references

Keywords: pca

Keywords: csbcbook

Keywords: kernel-theory

Keywords: bio

In this paper a method of estimating the parameters of a set of regression equations is reported which involves application of Aitken's generalized least-squares [1] to the whole system of equations. Under conditions generally encountered in practice, it is found that the regression coefficient estimators so obtained are at least asymptotically more efficient than those obtained by an equation-by-equation application of least squares. This gain in efficiency can be quite large if ïndependent" variables in different equations are not highly correlated and if disturbance terms in different equations are highly correlated. Further, tests of the hypothesis that all regression equation coefficient vectors are equal, based on "micro" and "macro" data, are described. If this hypothesis is accepted, there will be no aggregation bias. Finally, the estimation procedure and the "micro-test" for aggregation bias are applied in the analysis of annual investment data, 1935-1954, for two firms.

Keywords: 480-3, econometrics, sur

The central dogma of molecular biology deals with the detailed residue-by-residue transfer of sequential information. It states that such informatfon cannot be transferred from protein to either proteln or nucleic acid.

Keywords: csbcbook

Keywords: csbcbook

The probability of error in decoding an optimal convolutional code transmitted over a memoryless channel is bounded from above and below as a function of the constraint length of the code. For all but pathological channels the bounds are asymptotically (exponentially) tight for rates aboveR_0, the computational cutoff rate of sequential decoding. As a function of constraint length the performance of optimal convolutional codes is shown to be superior to that of block codes of the same length, the relative improvement increasing with rate. The upper bound is obtained for a specific probabilistic nonsequential decoding algorithm which is shown to be asymptotically optimum for rates aboveR_0and whose performance bears certain similarities to that of sequential decoding algorithms.

Keywords: criteria, model, selection

Universal coding is any asymptotically optimum method of block-to-block memoryless source coding for sources with unknown parameters. This paper considers noiseless coding for such sources, primarily in terms of variable-length coding, with performance measured as a function of the coding redundancy relative to the per-letter conditional source entropy given the unknown parameter. It is found that universal (i.e., zero redundancy) coding in a weighted sense is possible if and only if the per-letter average mutual information between the parameter space and the message space is zero. Universal coding is possible in a maximin sense if and only if the channel capacity between the two spaces is zero. Universal coding is possible in a minimax sense if and only if a probability mass function exists, independent of the unknown parameter, for which the relative entropy of the known conditional-probability mass-function is zero. Several examples are given to illustrate the ideas. Particular attention is given to sources that are stationary and ergodic for any fixed parameter although the whole ensemble is not. For such sources, weighted universal codes always exist if the alphabet is finite, or more generally if the entropy is finite. Minimax universal codes result if an additional entropy stability constraint is applied. A discussion of fixed-rate universal coding is also given briefly with performance measured by a probability of error.

Keywords: information-theory universal-coding

The problem of estimating the dimensionality of a model occurs in various forms in applied statistics. There is estimating the number of factor in factor analysis, estimating the degree of a polynomial describing the data, selecting the variables to

Keywords: conf, Akaike Information Criterion, criteria, AIC, model, modelling, parameters, complexity, overfitting, c1973, c197x, c19xx

Keywords: breastcancer

Keywords: Animals, Brain, Dihydroxyphenylalanine, Dose-Response Relationship, Drug, Drug Synergism, Indenes, Mathematics, Mice, Monoamine Oxidase Inhibitors, Naphthalenes, Oxidation-Reduction, Oxotremorine, Reserpine, Structure-Activity Relationship, Tryptamines, 4830537

Keywords: Computers; Great Britain; Information Systems; Japan; Protein Conformation; Proteins; United States

A new method of constructing a universal sequence of block codes for coding a class of ergodic sources is given. With this method, a weakly universal sequence of codes is constructed for variable-rate noise. less coding and for fixed- and variable-rate coding with respect to a fidelity criterion. In this way a unified approach to weak universal block source coding is obtained. For the noiseless variable-rate coding and the fixed-rate coding with respect to fidelity criterion, the assumptions made on the alphabets, distortion measures, and class of sources are both necessary and sufficient. For fixed-rate coding with respect to a fidelity criterion, the sample distortion of the universal code sequence converges inL^lnorm for each source to the optimum distortion for that source. For both variable-rate noiseless coding and variable-rate coding with respect to a fidelity criterion, the sample rate of the universal code sequence converges inL^1norm for each source to the optimum rate for that source. Using this fact, a universal sequence of codes for fixed-rate noiseless coding is obtained. Some applications to stationary nonergodic sources are also considered. The results of Davisson, Ziv, Neuhoff, Gray, Pursley, and Mackenthun are extended.

Keywords: universal-coding information-theory

Since 1971 conjoint analysis has been applied to a wide variety of problems in consumer research. This paper discusses various issues involved in implementing conjoint analysis and describes some new technical developments and application areas for the methodology.

Keywords: conjoint_analysis

Keywords: chemoinformatics

Keywords: chemoinformatics

Keywords: information-theory source-coding

Universal coding theory is surveyed from the viewpoint of the interplay between delay and redundancy. The price for universality turns out to be acceptably small.

Keywords: information-theory source-coding

A new paradigm, Random Sample Consensus (RANSAC), for fitting a model to experimental data is introduced. RANSAC is capable of interpreting/smoothing data containing a significant percentage of gross errors, and is thus ideally suited for applications in automated image analysis where interpretation is based on the data provided by error-prone feature detectors. A major portion of this paper describes the application of RANSAC to the Location Determination Problem (LDP): Given an image depicting a set of landmarks with know locations, determine that point in space from which the image was obtained. In response to a RANSAC requirement, new results are derived on the minimum number of landmarks needed to obtain a solution, and algorithms are presented for computing these minimum-landmark solutions in closed form. These results provide the basis for an automatic system that can solve the LDP under difficult viewing.

This paper is concerned with the inexact matching of attributed, relational graphs for structural pattern recognition. The matching procedure is based on a state space search utilizing heuristic information. Some experimental results are reported.

Keywords: graph, matching

Keywords: Mass Spectrometry; Peptides; Terminology as Topic

A connection between universal codes and the problems of prediction and statistical estimation is established. A known lower bound for the mean length of universal codes is sharpened and generalized, and optimum universal codes constructed. The bound is defined to give the information in strings relative to the considered class of processes. The earlier derived minimum description length criterion for estimation of parameters, including their number, is given a fundamental information, theoretic justification by showing that its estimators achieve the information in the strings. It is also shown that one cannot do prediction in Gaussian autoregressive moving average (ARMA) processes below a bound, which is determined by the information in the data.

Keywords: information-theory universal-coding

LetX_i_i=1^inftybe a sequence of independent Bernoulli random variables with probabilitypthatX_i = 1and probabilityq=1-pthatX_i = 0for alli geq 1. Time-invariant finite-memory (i.e., finite-state) estimation procedures for the parameter p are considered which takeX_1, cdotsas an input sequence. In particular, an n-state deterministic estimation procedure is described which can estimate p with mean-square errorO(log n/n)and ann-state probabilistic estimation procedure which can estimatepwith mean-square errorO(1/n). It is proved that theO(1/n)bound is optimal to within a constant factor. In addition, it is shown that linear estimation procedures are just as powerful (up to the measure of mean-square error) as arbitrary estimation procedures. The proofs are based on an analog of the well-known matrix tree theorem that is called the Markov chain tree theorem.

Thermodynamic parameters for prediction of RNA duplex stability are reported. One parameter for duplex initiation and 10 parameters for helix propagation are derived from enthalpy and free-energy changes for helix formation by 45 RNA oligonucleotide duplexes. The oligomer sequences were chosen to maximize reliability of secondary structure predictions. Each of the 10 nearest-neighbor sequences is well-represented among the 45 oligonucleotides, and the sequences were chosen to minimize experimental errors in delta GO at 37 degrees C. These parameters predict melting temperatures of most oligonucleotide duplexes within 5 degrees C. This is about as good as can be expected from the nearest-neighbor model. Free-energy changes for helix propagation at dangling ends, terminal mismatches, and internal G X U mismatches, and free-energy changes for helix initiation at hairpin loops, internal loops, or internal bulges are also tabulated.

The Maximum Entropy (ME) and Maximum Likelihood (ML) criteria are the bases for two approaches to statistical inference problems. A new criterion, called the Minimum Description Length (MDL), has been recently introduced. This criterion generalizes the ML method so it can be applied to more general situations, e.g., when the number of parameters is unknown. It is shown that ME is also a special case of the MDL criterion; maximizing the entropy subject to some constraints on the underlying probability function is identical to minimizing the code length required to represent all possible i.i.d, realizations of the random variable such that the sample frequencies (or histogram) satisfy those given constraints.

Keywords: information-theory

An approximate solution to the weighted-graph-matching problem is discussed for both undirected and directed graphs. The weighted-graph-matching problem is that of finding the optimum matching between two weighted graphs, which are graphs with weights at each arc. The proposed method uses an analytic instead of a combinatorial or iterative approach to the optimum matching problem. Using the eigendecompositions of the adjacency matrices (in the case of the undirected-graph-matching problem) or Hermitian matrices derived from the adjacency matrices (in the case of the directed-graph-matching problem), a matching close to the optimum can be found efficiently when the graphs are sufficiently close to each other. Simulation results are given to evaluate the performance of the proposed method.

Fifteen RNA hairpins that share the same stem sequence and have homopolymer loops of A, C and U residues which vary in length from three to nine nucleotides were synthesized and their thermal stabilities determined. Tm varies as a function of loop size but is almost independent of loop composition. Loops of four or five nucleotides are found to be the most stable loop size. This is consistent with the observation that four-membered loops are the most prevalent loop size in 16S-like RNAs. The contribution of each loop to hairpin stability was calculated by subtracting the known contribution of the helical stem. These data should be useful for predicting the stability of other hairpins.

A test sequence is used to select the best rule from a class of discrimination rules defined in terms of the training sequence. The Vapnik-Chervonenkis and related inequalities are used to obtain distribution-free bounds on the difference between the probability of error of the selected rule and the probability of error of the best rule in the given class. The bounds are used to prove the consistency and asymptotic optimality for several popular classes, including linear discriminators, nearest-neighbor rules, kernel-based rules, histogram rules, binary tree classifiers, and Fourier series classifiers. In particular, the method can be used to choose the smoothing parameter in kernel-based rules, to choose k in the k-nearest neighbor rule, and to choose between parametric and nonparametric rules

Keywords: information-theory

It is shown that each bit of information at most doubles the resulting wealth in the general stock-market setup. This information bound on the growth of wealth is actually attained for certain probability distributions on the market investigated by J. Kelly (1956). The bound is shown to be a special case of the result that the increase in exponential growth of wealth achieved with true knowledge of the stock market distribution F over that achieved with incorrect knowledge G is bounded above by the entropy of F relative to G

Keywords: information-theory

An algorithm and a computer program have been prepared for determining RNA secondary structures within any prescribed increment of the computed global minimum free energy. The mathematical problem of determining how well defined a minimum energy folding is can now be solved. All predicted base pairs that can participate in suboptimal structures may be displayed and analyzed graphically. Representative suboptimal foldings are generated by selecting these base pairs one at a time and computing the best foldings that contain them. A distance criterion that ensures that no two structures are "too close" is used to avoid multiple generation of similar structures. Thermodynamic parameters, including free-energy increments for single-base stacking at the ends of helices and for terminal mismatched pairs in interior and hairpin loops, are incorporated into the underlying folding model of the above algorithm.

Keywords: sirna

Theorems concerning the entropy of a stationary ergodic information source are derived and used to obtain insight into the workings of certain data-compression coding schemes, in particular the Lempel-Siv data compression algorithm

Keywords: information-theory

A novel universal data compression algorithm is described. This algorithm encodes L source symbols at a time. An upper limit for the number of bits per source symbol is given for the class of binary stationary sources. In the author's analysis, a property of repetition times turns out to be of crucial importance

Protein-protein interactions between two proteins have generally been studied using biochemical techniques such as crosslinking, co-immunoprecipitation and co-fractionation by chromatography. We have generated a novel genetic system to study these interactions by taking advantage of the properties of the GAL4 protein of the yeast Saccharomyces cerevisiae. This protein is a transcriptional activator required for the expression of genes encoding enzymes of galactose utilization. It consists of two separable and functionally essential domains: an N-terminal domain which binds to specific DNA sequences (UASG); and a C-terminal domain containing acidic regions, which is necessary to activate transcription. We have generated a system of two hybrid proteins containing parts of GAL4: the GAL4 DNA-binding domain fused to a protein 'X' and a GAL4 activating region fused to a protein 'Y'. If X and Y can form a protein-protein complex and reconstitute proximity of the GAL4 domains, transcription of a gene regulated by UASG occurs. We have tested this system using two yeast proteins that are known to interact-SNF1 and SNF4. High transcriptional activity is obtained only when both hybrids are present in a cell. This system may be applicable as a general method to identify proteins that interact with a known protein by the use of a simple galactose selection.

The Complex Carbohydrate Structure Database (CCSD) and CarbBank, an IBM PC/AT (or compatible) database management system, were created to provide an information system to meet the needs of people interested in carbohydrate science. The CCSD, which presently contains more than 2000 citations, is expected to double in size in the next two years and to include, soon thereafter, all of the published structures of carbohydrates larger than disaccharides.

Keywords: Carbohydrate Sequence, Carbohydrates, Databases, Factual, Information Systems, Molecular Structure, 2623761

The decomposition of deformations by principal warps is demonstrated. The method is extended to deal with curving edges between landmarks. This formulation is related to other applications of splines current in computer vision. How they might aid in the extraction of features for analysis, comparison, and diagnosis of biological and medical images in indicated

An application of the neural network to quantitative structure-activity relationship (QSAR) analysis has been studied. The new method was compared with the linear multiregression analysis in various ways. It was found that the neural network can be a potential tool in the routine work of QSAR analysis. The mathematical relationship of operation between the neural network and the multiregression analysis was described. It was shown that the neural network can exceed the level of the linear multiregression analysis.

Keywords: Animals, Azirines, Benzodiazepines, Carbazilquinone, Comparative Study, Nerve Net, Nervous System, Regression Analysis, Structure-Activity Relationship, 2202830

In the absence of knowledge of the true density function, Bayesian models take the joint density function for a sequence of n random variables to be an average of densities with respect to a prior. The authors examine the relative entropy distance Dn between the true density and the Bayesian density and show that the asymptotic distance is (d/2)(log n)+c, where d is the dimension of the parameter vector. Therefore, the relative entropy rate Dn/n converges to zero at rate (log n)/n. The constant c, which the authors explicitly identify, depends only on the prior density function and the Fisher information matrix evaluated at the true parameter value. Consequences are given for density estimation, universal data compression, composite hypothesis testing, and stock-market portfolio selection

Keywords: information-theory

This paper reports a comparison of several methods for measuring the degree of similarity between pairs of 3-D chemical structures that are represented by inter-atomic distance matrices. The methods that have been tested use the distance information in very different ways and have very different computational requirements. Experiments with 10 small datasets, for which both structural and biological activity data are available, suggest that the most cost-effective technique is based on a mapping procedure that tries to match pairs of atoms, one from each of the molecules that are being compared, that have neighbouring atoms at approximately the same distances.

Keywords: Algorithms, Binding Sites, Chemical, Chemistry, Comparative Study, Computer Simulation, Databases, Factual, Macromolecular Substances, Models, Molecular Conformation, Molecular Structure, Non-U.S. Gov't, Physical, Protein Conformation, Protein Structure, Proteins, Research Support, Structure-Activity Relationship, Tertiary, 1770381

Morphological assessment of the degree of differentiation has been shown in numerous studies to provide useful prognostic information in breast cancer, but until recently histological grading has not been accepted as a routine procedure, mainly because of perceived problems with reproducibility and consistency. In the Nottingham/Tenovus Primary Breast Cancer Study the most commonly used method, described by Bloom & Richardson, has been modified in order to make the criteria more objective. The revised technique involves semiquantitative evaluation of three morphological features-the percentage of tubule formation, the degree of nuclear pleomorphism and an accurate mitotic count using a defined field area. A numerical scoring system is used and the overall grade is derived from a summation of individual scores for the three variables: three grades of differentiation are used. Since 1973, over 2200 patients with primary operable breast cancer have been entered into a study of multiple prognostic factors. Histological grade, assessed in 1831 patients, shows a very strong correlation with prognosis; patients with grade I tumours have a significantly better survival than those with grade II and III tumours (P less than 0.0001). These results demonstrate that this method for histological grading provides important prognostic information and, if the grading protocol is followed consistently, reproducible results can be obtained. Histological grade forms part of the multifactorial Nottingham prognostic index, together with tumour size and lymph node stage, which is used to stratify individual patients for appropriate therapy.

Keywords: csbcbook, csbcbook-ch3

Keywords: chemoinformatics

The authors introduce an index of resolvability that is proved to bound the rate of convergence of minimum complexity density estimators as well as the information-theoretic redundancy of the corresponding total description length. The results on the index of resolvability demonstrate the statistical effectiveness of the minimum description-length principle as a method of inference. The minimum complexity estimator converges to true density nearly as fast as an estimator based on prior knowledge of the true subclass of densities. Interpretations and basic properties of minimum complexity estimators are discussed. Some regression and classification problems that can be examined from the minimum description-length framework are considered

Keywords: information-theory

Back propagation neural networks is a new technology useful for modeling nonlinear functions of several variables. This paper explores their applications in the field of quantitative structure-activity relationships. In particular, their ability to fit biological activity surfaces, predict activity, and determine the "functional forms" of its dependence on physical properties is compared to well-established methods in the field. A dataset of 256 5-phenyl-3,4-diamino-6,6-dimethyldihydrotriazines that inhibit dihydrofolate reductase enzyme is used as a basis for comparison. It is found that neural networks lead to enhanced surface fits and predictions relative to standard regression methods. Moreover, they circumvent the need for ad hoc indicator variables, which account for a significant part of the variance in linear regression models. Additionally, they lead to the elucidation of nonlinear and "cross-products" effects that correspond to trade-offs between physical properties in their effect on biological activity. This is the first demonstration of the latter two findings. On the other hand, due to the complexity of the resulting models, an understanding of the local, but not the global, structure-activity relationships is possible. The latter must await further developments. Furthermore, the longer computational time required to train the networks is somewhat inconveniencing, although not restrictive.

Keywords: Animals, Carcinoma 256, Cultured, Experimental, Folic Acid Antagonists, Leukemia, Neural Pathways, Neurons, Regression Analysis, Structure-Activity Relationship, Tumor Cells, Walker, 1895302

It is shown that the modified Lawrence algorithm is universal over the class of binary memoryless sources and that the rate converges asymptotically optimally fast to the source entropy. It is proven that no codes exist that have a better asymptotic performance. The asymptotic bounds show that universal variable-to-fixed-length codes can have a significantly lower redundancy than universal fixed-to-variable-length codes with the same number of codewords

Keywords: information-theory source-coding

The peptides presented by major histocompatibility complex class I molecules adhere to strict rules concerning peptide length and occupancy by certain amino acid residues at anchor positions. Peptides presented by HLA-A*0201 molecules, for example, are generally nonapeptides requiring Leu or Met at position 2 and an aliphatic residue, predominantly Val, at position 9. A closely related molecule, HLA-A*0205, differing from the former at four amino acid residues, has a related but substantially different peptide motif. A*0205-presented peptides are still nonapeptides, and position 9 is still aliphatic, although it is preferentially occupied by Leu instead of Val. Position 2 not only allows aliphatic residues but also polar ones. Occupancy at position 6, considered as an auxiliary anchor in A*0201, as well as non-anchor residues at positions 3, 4, and 8 are relatively well conserved between the two peptide motifs. Thus, although a number of the T cell epitopes presented by the two HLA-A2 forms is expected to be identical, a considerable number of epitopes should be different.

Keywords: immunoinformatics

An algorithm is presented for the accurate and rapid generation of multiple protein sequence alignments from tertiary structure comparisons. A preliminary multiple sequence alignment is performed using sequence information, which then determines an initial superposition of the structures. A structure comparison algorithm is applied to all pairs of proteins in the superimposed set and a similarity tree calculated. Multiple sequence alignments are then generated by following the tree from the branches to the root. At each branchpoint of the tree, a structure-based sequence alignment and coordinate transformations are output, with the multiple alignment of all structures output at the root. The algorithm encoded in STAMP (STructural Alignment of Multiple Proteins) is shown to give alignments in good agreement with published structural accounts within the dehydrogenase fold domains, globins, and serine proteinases. In order to reduce the need for visual verification, two similarity indices are introduced to determine the quality of each generated structural alignment. Sc quantifies the global structural similarity between pairs or groups of proteins, whereas Pij' provides a normalized measure of the confidence in the alignment of each residue. STAMP alignments have the quality of each alignment characterized by Sc and Pij' values and thus provide a reproducible resource for studies of residue conservation within structural motifs.

Keywords: Algorithms; Amino Acid Sequence; Animals; Confidence Intervals; Globins; Humans; Molecular Sequence Data; Protein Structure, Tertiary; Sequence Alignment; Sequence Homology, Amino Acid; Serine Endopeptidases; Software

A constrained optimization type of numerical algorithm for removing noise from images is presented. The total variation of the image is minimized subject to constraints involvingthe statistics of the noise. The constraints are imposed using Lagrange multipliers. The solution is obtained using the gradient-projection method. This amounts to solving a time dependent partial differential equation on a manifold determined by the constraints. As t- 0othe solution converges to a steady state which is the denoised image. The numerical algorithm is simple and relatively fast. The results appear to be state-of-the-art for very noisy images. The method is noninvasive, yielding sharp edges in the image. The technique could be interpreted as a first step of moving each level set of the the level set divided by the magnitude of the gradient of the the constraint set.

Keywords: segmentation

The results by P. Hall and E.J. Hannan (1988) on optimization of histogram density estimators with equal bin widths by minimization of the stochastic complexity are extended and sharpened in two separate ways. As the first contribution, two generalized histogram estimators are constructed. The first has unequal bin widths which, together with the number of the bins, are determined by minimization of the stochastic complexity using dynamic programming. The other estimator consists of a mixture of equal bin width estimators, each of which is defined by the associated stochastic complexity. As the main contribution in the present work, two theorems are proved, which together extend the universal coding theorems to a large class of data generating densities. The first gives an asymptotic upper bound for the code redundancy in the order of magnitude, achieved with a special predictive type of histogram estimator, which sharpens a related bound. The second theorem states that this bound cannot be improved upon by any code whatsoever

Keywords: information-theory

The machine learning program GOLEM from the field of inductive logic programming was applied to the drug design problem of modeling structure-activity relationships. The training data for the program were 44 trimethoprim analogues and their observed inhibition of Escherichia coli dihydrofolate reductase. A further 11 compounds were used as unseen test data. GOLEM obtained rules that were statistically more accurate on the training data and also better on the test data than a Hansch linear regression model. Importantly machine learning yields understandable rules that characterized the chemistry of favored inhibitors in terms of polarity, flexibility, and hydrogen-bonding character. These rules agree with the stereochemistry of the interaction observed crystallographically.

Keywords: Algorithms, Artificial Intelligence, Drug Design, Escherichia coli, Folic Acid Antagonists, Molecular Structure, Mutagens, Nitroso Compounds, Non-U.S. Gov't, Research Support, Structure-Activity Relationship, Trimethoprim, 1454814

Comparative genomic hybridization produces a map of DNA sequence copy number as a function of chromosomal location throughout the entire genome. Differentially labeled test DNA and normal reference DNA are hybridized simultaneously to normal chromosome spreads. The hybridization is detected with two different fluorochromes. Regions of gain or loss of DNA sequences, such as deletions, duplications, or amplifications, are seen as changes in the ratio of the intensities of the two fluorochromes along the target chromosomes. Analysis of tumor cell lines and primary bladder tumors identified 16 different regions of amplification, many in loci not previously known to be amplified.

Keywords: csbcbook, csbcbook-ch2, cgh

Methods for alignment of protein sequences typically measure similarity by using a substitution matrix with scores for all possible exchanges of one amino acid with another. The most widely used matrices are based on the Dayhoff model of evolutionary rates. Using a different approach, we have derived substitution matrices from about 2000 blocks of aligned sequence segments characterizing more than 500 groups of related proteins. This led to marked improvements in alignments and in searches using queries from each of the groups.

Keywords: Algorithms; Amino Acid Sequence; Animals; Caenorhabditis elegans; Drosophila; Lod Score; Mathematics; Molecular Sequence Data; Probability; Proteins; Sequence Homology, Amino Acid; Software

The problem of predicting the next outcome of an individual binary sequence using finite memory is considered. The finite-state predictability of an infinite sequence is defined as the minimum fraction of prediction errors that can be made by any finite-state (FS) predictor. It is proven that this FS predictability can be achieved by universal sequential prediction schemes. An efficient prediction procedure based on the incremental parsing procedure of the Lempel-Ziv data compression algorithm is shown to achieve asymptotically the FS predictability. Some relations between compressibility and predictability are discussed, and the predictability is proposed as an additional measure of the complexity of a sequence

Keywords: information-theory universal-coding

The histological tumour type determined by current criteria has been investigated in a consecutive series of 1621 women with primary operable breast carcinoma, presenting between 1973 and 1987. All women underwent definitive surgery with node biopsy and none received adjuvant systemic therapy. Special types, tubular, invasive cribriform and mucinous, with a very favourable prognosis can be identified. A common type of tumour recognized by our group and designated tubular mixed carcinoma is shown to be prognostically distinct from carcinomas of no special type; it has a characteristic histological appearance and is the third most common type in this series. Analysis of subtypes of lobular carcinoma confirms differing prognoses. The classical, tubulo-lobular and lobular mixed types are associated with a better prognosis than carcinomas of no special type; this is not so for the solid variant. Tubulo-lobular carcinoma in particular has an extremely good prognosis similar to tumours included in the 'special type' category above. Neither medullary carcinoma nor atypical medullary carcinoma are found to carry a survival advantage over carcinomas of no special type. The results confirm that histological typing of human breast carcinoma can provide useful prognostic information.

Keywords: csbcbook, csbcbook-ch3

Keywords: glycans

The problem of the nonparametric estimation of a probability distribution is considered from three viewpoints: the consistency in total variation, the consistency in information divergence, and consistency in reversed-order information divergence. These types of consistencies are relatively strong criteria of convergence, and a probability distribution cannot be consistently estimated in either type of convergence without any restrictions on the class of probability distributions allowed. Histogram-based estimators of distribution are presented which, under certain conditions, converge in total variation, in information divergence, and in reversed-order information divergence to the unknown probability distribution. Some a priori information about the true probability distribution is assumed in each case. As the concept of consistency in information divergence is stronger than that of convergence in total variation, additional assumptions are imposed in the cases of informational divergences

Recently, methods have been developed in the field of Artificial Intelligence (AI), specifically in the expert systems area using rule-induction, designed to extract rules from data. We have applied these methods to the analysis of molecular series with the objective of generating rules which are predictive and reliable. The input to rule-induction consists of a number of examples with known outcomes (a training set) and the output is a tree-structured series of rules. Unlike most other analysis methods, the results of the analysis are in the form of simple statements which can be easily interpreted. These are readily applied to new data giving both a classification and a probability of correctness. Rule-induction has been applied to in-house generated and published QSAR datasets and the methodology, application and results of these analyses are discussed. The results imply that in some cases it would be advantageous to use rule-induction as a complementary technique in addition to conventional statistical and pattern-recognition methods.

Keywords: Algorithms, Anticonvulsants, Antimalarials, Artificial Intelligence, Cardiotonic Agents, Software, Structure-Activity Relationship, gamma-Aminobutyric Acid, 1403028

Recommendations of the Nomenclature Committee of the International Union of Biochemistry and Molecular Biology on the Nomenclature and Classification of Enzymes.

Keywords: gene-ontology, ontology

The expected redundancy per symbol of an n-block prefix code Cn on a source ? measures how far the code is from being optimal for that source. The existence of sequences of codes with expected redundancy per symbol of O((log n)/n) for `nice' classes of sources, such as Markov sources of a given order, is well known. It is shown that some restriction on the class of processes is necessary in order to obtain such redundancy bounds, for there is no universal redundancy rate for any sequence of prefix codes on the class of all ergodic sources

Keywords: information-theory

Keywords: chemoinformatics

Some new ways of defining the entropy of a process by observing a single typical output sequence as well as a new kind of Shannon-McMillan-Breiman theorem are presented. This provides a new and conceptually very simple ways of estimating the entropy of an ergodic stationary source as well as new insight into the workings of such well-known data compression schemes as the Lempel-Ziv algorithm

Keywords: information-theory

A loss function, or objective function, is a function used to compare parameters when fitting a model to data. The loss function gives a distance between the model output and the desired output. Two common examples are the squared-error loss function and the cross entropy loss function. Minimizing the mean-square error loss function is equivalent to minimizing the mean square difference between the model output and the expected value of the output given a particular input. This property of minimization to the expected value is formalized as P-admissibility. The necessary and sufficient conditions for P-admissibility, leading to a parametric description of all P-admissible loss functions, are found. In particular, it is shown that two of the simplest members of this class of functions are the squared error and the cross entropy loss functions. One application of this work is in the choice of a loss function for training neural networks to provide probability estimates

Sequential decision algorithms are investigated in relation to a family of additive performance criteria for individual data sequences. Simple universal sequential schemes are known, under certain conditions, to approach optimality uniformly as fast as n-1 log n, where n is the sample size. For the case of finite-alphabet observations, the class of schemes that can be implemented by finite-state machines (FSMs) is studied. It is shown that Markovian machines with sufficiently long memory exist, which are asymptotically nearly as good as any given deterministic or randomized FSM for the purpose of sequential decision. For the continuous-valued observation case, a useful class of parametric schemes is discussed with special attention to the recursive least squares algorithm

Keywords: information-theory source-coding

The authors introduce an algorithm, called matching pursuit, that decomposes any signal into a linear expansion of waveforms that are selected from a redundant dictionary of functions. These waveforms are chosen in order to best match the signal structures. Matching pursuits are general procedures to compute adaptive signal representations. With a dictionary of Gabor functions a matching pursuit defines an adaptive time-frequency transform. They derive a signal energy distribution in the time-frequency plane, which does not include interference terms, unlike Wigner and Cohen class distributions. A matching pursuit isolates the signal structures that are coherent with respect to a given dictionary. An application to pattern extraction from noisy signals is described. They compare a matching pursuit decomposition with a signal expansion over an optimized wavepacket orthonormal basis, selected with the algorithm of Coifman and Wickerhauser see (IEEE Trans. Informat. Theory, vol. 38, Mar. 1992)

Keywords: pursuit

In statistical pattern recognition, a classifier is called universally consistent if its error probability converges to the Bayes-risk as the size of the training data grows for all possible distributions of the random variable pair of the observation vector and its class. It is proven that if a one-layered neural network with properly chosen number of nodes is trained to minimize the empirical risk on the training data, then a universally consistent classifier results. It is shown that the exponent in the rate of convergence does not depend on the dimension if certain smoothness conditions on the distribution are satisfied. That is, this class of universally consistent classifiers does not suffer from the curse of dimensionality. A training algorithm is presented that finds the optimal set of parameters in polynomial time if the number of nodes and the space dimension is fixed and the amount of training data grows

Universal predictions of the next outcome of a binary sequence drawn from a Markov source with unknown parameters is considered. For a given source, the predictability is defined as the least attainable expected fraction of prediction errors. A lower bound is derived on the maximum rate at which the predictability is asymptotically approached uniformly over all sources in the Markov class. This bound is achieved by a simple majority predictor. For Bernoulli sources, bounds on the large deviations performance are investigated. A lower bound is derived for the probability that the fraction of errors will exceed the predictability by a prescribed amount ?>0. This bound is achieved by the same predictor if ? is sufficiently small

Keywords: information-theory source-coding

Approximation properties of a class of artificial neural networks are established. It is shown that feedforward networks with one layer of sigmoidal nonlinearities achieve integrated squared error of order O (1/n), where n is the number of nodes. The approximated function is assumed to have a bound on the first moment of the magnitude distribution of the Fourier transform. The nonlinear parameters associated with the sigmoidal nodes, as well as the parameters of linear combination, are adjusted in the approximation. In contrast, it is shown that for series expansions with n terms, in which only the parameters of linear combination are adjusted, the integrated squared approximation error cannot be made smaller than order 1/n2d/ uniformly for functions satisfying the same smoothness assumption, where d is the dimension of the input to the function. For the class of functions examined, the approximation rate and the parsimony of the parameterization of the networks are shown to be advantageous in high-dimensional settings

Keywords: information-theory

The problem of sequential probability assignment for individual sequences is investigated. The authors compare the probabilities assigned by any sequential scheme to the performance of the best ?batch? scheme (model) in some class. For the class of finite-state schemes and other related families, they derive a deterministic performance bound, analogous to the classical (probabilistic) minimum description length (MDL) bound. It holds for ?most? sequences, similarly to the probabilistic setting, where the bound holds for ?most? sources in a class. It is shown that the bound can be attained both pointwise and sequentially for any model family in the reference class and without any prior knowledge of its order. This is achieved by a universal scheme based on a mixing approach. The bound and its sequential achievability establish a completely deterministic significance to the concept of predictive MDL

Keywords: information-theory

We have identified a conserved family of genes related to Drosophila eag, which encodes a distinct type of voltage-activated K+ channel. Three related genes were recovered in screens of cDNA libraries from Drosophila, mouse, and human tissues. One gene is the mouse counterpart of eag; the other two represent additional subfamilies. The human gene maps to chromosome 7. Family members share at least 47% amino acid identity in their hydrophobic cores and all contain a segment homologous to a cyclic nucleotide-binding domain. Sequence comparisons indicate that members of this family are most closely related to vertebrate cyclic nucleotide-gated cation channels and plant inward-rectifying K+ channels. The existence of another family of K+ channel structural genes further extends the known diversity of K+ channels and has important implications for the structure, function, and evolution of the superfamily of voltage-sensitive ion channels.

The relationship between binding affinity for HLA class I molecules and immunogenicity of discrete peptide epitopes has been analyzed in two different experimental approaches. In the first approach, the immunogenicity of potential epitopes ranging in MHC binding affinity over a 10,000-fold range was analyzed in HLA-A*0201 transgenic mice. In the second approach, the antigenicity of approximately 100 different hepatitis B virus (HBV)-derived potential epitopes, all carrying A*0201 binding motifs, was assessed by using PBL of acute hepatitis patients. In both cases, it was found that an affinity threshold of approximately 500 nM (preferably 50 nM or less) apparently determines the capacity of a peptide epitope to elicit a CTL response. These data correlate well with class I binding affinity measurements of either naturally processed peptides or previously described T cell epitopes. Taken together, these data have important implications for the selection of epitopes for peptide-based vaccines, and also formally demonstrate the crucial role of determinant selection in the shaping of T cell responses. Because in most (but not all) cases, high affinity peptides seem to be immunogenic, our data also suggest that holes in the functional T cell repertoire, if they exist, may be relatively rare.

Keywords: Amino Acid Sequence; Animals; Cell Line; Cytotoxicity Tests, Immunologic; Epitopes; HLA-A Antigens; Hepatitis B; Hepatitis B Antigens; Humans; Mice; Mice, Transgenic; Molecular Sequence Data; Peptides; Protein Binding; T-Lymphocytes, Cytotoxic

A method to predict the relative binding strengths of all possible nonapeptides to the MHC class I molecule HLA-A2 has been developed based on experimental peptide binding data. These data indicate that, for most peptides, each side-chain of the peptide contributes a certain amount to the stability of the HLA-A2 complex that is independent of the sequence of the peptide. To quantify these contributions, the binding data from a set of 154 peptides were combined together to generate a table containing 180 coefficients (20 amino acids x 9 positions), each of which represents the contribution of one particular amino acid residue at a specified position within the peptide to binding to HLA-A2. Eighty peptides formed stable HLA-A2 complexes, as assessed by measuring the rate of dissociation of beta 2m. The remaining 74 peptides formed complexes that had a half-life of beta 2m dissociation of less than 5 min at 37 degrees C, or did not bind to HLA-A2, and were included because they could be used to constrain the values of some of the coefficients. The "theoretical" binding stability (calculated by multiplying together the corresponding coefficients) matched the experimental binding stability to within a factor of 5. The coefficients were then used to calculate the theoretical binding stability for all the previously identified self or antigenic nonamer peptides known to bind to HLA-A2. The binding stability for all other nonamer peptides that could be generated from the proteins from which these peptides were derived was also predicted. In every case, the previously described HLA-A2 binding peptides were ranked in the top 2% of all possible nonamers for each source protein. Therefore, most biologically relevant nonamer peptides should be identifiable using the table of coefficients. We conclude that the side-chains of most nonamer peptides to the first approximation bind independently of one another to the HLA-A2 molecule.

Keywords: immunoinformatics

Shows that a discrete infinite distribution with finite entropy cannot be estimated consistently in information divergence. As a corollary the authors show that there is no universal source code for an infinite source alphabet over the class of all discrete memoryless sources with finite entropy

Keywords: information-theory

A new criterion is proposed for the evaluation of variable selection procedures in multiple regression. This criterion, which we call the risk inflation, is based on an adjustment to the risk. Essentially, the risk inflation is the maximum increase in risk due to selecting rather than knowing the "correct" predictors. A new variable selection procedure is obtained which, in the case of orthogonal predictors, substantially improves on AIC, Cp and BIC and is close to optimal. In contrast to AIC, Cp and BIC which use dimensionality penalties of 2, 2 and log n, respectively, this new procedure uses a penalty 2 log p, where p is the number of available predictors. For the case of nonorthogonal predictors, bounds for the optimal penalty are obtained.

Keywords: mdl, regression

The relation between the entropy of a discrete random variable and the minimum attainable probability of error made in guessing its value is examined. While Fano's inequality provides a tight lower bound on the error probability in terms of the entropy, the present authors derive a converse result-a tight upper bound on the minimal error probability in terms of the entropy. Both bounds are sharp, and can draw a relation, as well, between the error probability for the maximum a posteriori (MAP) rule, and the conditional entropy (equivocation), which is a useful uncertainty measure in several applications. Combining this relation and the classical channel coding theorem, the authors present a channel coding theorem for the equivocation which, unlike the channel coding theorem for error probability, is meaningful at all rates. This theorem is proved directly for DMCs, and from this proof it is further concluded that for R⩾C the equivocation achieves its minimal value of R-C at the rate of n1/2 where n is the block length

Keywords: information-theory

We provide a rigorous proof that Jeffreys' prior asymptotically maximizes Shannon's mutual information between a sample of size n and the parameter. This was conjectured by Bernardo (1979) and, despite the absence of a proof, forms the basis of the reference prior method in Bayesian statistical analysis. Our proof rests on an examination of large sample decision theoretic properties associated with the relative entropy or the Kullback?Leibler distance between probability density functions for independent and identically distributed random variables. For smooth finite-dimensional parametric families we derive an asymptotic expression for the minimax risk and for the related maximin risk. As a result, we show that, among continuous positive priors, Jeffreys' prior uniquely achieves the asymptotic maximin value. In the discrete parameter case we show that, asymptotically, the Bayes risk reduces to the entropy of the prior so that the reference prior is seen to be the maximum entropy prior. We identify the physical significance of the risks by giving two information-theoretic interpretations in terms of probabilistic coding.

Keywords: information-theory

The algorithm described in this paper discovers one or more motifs in a collection of DNA or protein sequences by using the technique of expectation maximization to fit a two-component finite mixture model to the set of sequences. Multiple motifs are found by fitting a mixture model to the data, probabilistically erasing the occurrences of the motif thus found, and repeating the process to find successive motifs. The algorithm requires only a set of unaligned sequences and a number specifying the width of the motifs as input. It returns a model of each motif and a threshold which together can be used as a Bayes-optimal classifier for searching for occurrences of the motif in other databases. The algorithm estimates how many times each motif occurs in each sequence in the dataset and outputs an alignment of the occurrences of the motif. The algorithm is capable of discovering several different motifs with differing numbers of occurrences in a single dataset.

Combines optimization and ergodic theory to characterize the optimum long-run average performance that can be asymptotically attained by nonanticipating sequential decisions. Let Xt be a stationary ergodic process, and suppose an action bt must be selected in a space ℬ with knowledge of the t-past (X0, ···, Xt-1) at the beginning of every period t⩾0. Action bt will incur a loss l(bt, Xt) at the end of period t when the random variable Xt is revealed. The author proves under mild integrability conditions that the optimum strategy is to select actions that minimize the conditional expected loss given the currently available information at each step. The minimum long-run average loss per decision can be approached arbitrarily closely by strategies that are finite-order Markov, and under certain continuity conditions, it is equal to the minimum expected loss given the infinite past. If the loss l(b, x) is bounded and continuous and if the space ℬ is compact, then the minimum can be asymptotically attained, even if the distribution of the process Xt is unknown a priori and must be learned from experience

Keywords: information-theory

Keywords: svms

Herein we describe the establishment of assays to measure peptide binding to purified HLA-B*0701, -B*0801, -B*2705, -B*3501-03, -B*5401, -Cw*0401, -Cw*0602, and -Cw*0702 molecules. The binding of known peptide epitopes or naturally processed peptides correlates well with HLA restriction or origin, underscoring the immunologic relevance of these assays. Analysis of the sequences of various HLA class I alleles suggested that alleles with peptide motifs characterized by proline in position 2 and aromatic or hydrophobic residues at their C-terminus shared key consensus residues at positions 9, 63, 66, 67, and 70 (B pocket) and residue 116 (F pocket). Prediction of the peptide-binding specificity of HLA-B*5401, on the basis of this consensus B and F pocket structure, verified this hypothesis and suggested that a relatively large family of HLA-B alleles (which we have defined as the HLA-B7-like supertype) may significantly overlap in peptide binding specificity. Availability of quantitative binding assays allowed verification that, indeed, many (25%) of the peptide ligands carrying proline in position 2 and hydrophobic/aromatic residues at the C-terminus (the B7-like supermotif) were capable of binding at least three of five HLA-B7-like supertype alleles. Identification of epitopes carrying the B7-like supermotif and binding to a family of alleles represented in over 40% of individuals from all major ethnic groups may be of considerable use in the design of peptide vaccines.

Keywords: Alleles; Amino Acid Sequence; Cell Line, Transformed; Consensus Sequence; Epitopes; Genes, MHC Class I; HLA-B Antigens; HLA-C Antigens; Humans; Molecular Sequence Data; Peptide Fragments; Protein Binding; Protein Structure, Tertiary; Structure-Activity Relationship; Substrate Specificity

A high-capacity system was developed to monitor the expression of many genes in parallel. Microarrays prepared by high-speed robotic printing of complementary DNAs on glass were used for quantitative expression measurements of the corresponding genes. Because of the small format and high density of the arrays, hybridization volumes of 2 microliters could be used that enabled detection of rare transcripts in probe mixtures derived from 2 micrograms of total cellular messenger RNA. Differential expression measurements of 45 Arabidopsis genes were made by means of simultaneous, two-color fluorescence hybridization.

Keywords: microarray

We present a new method for docking flexible peptides to class I Major-Histocompatibility-Complex (MHC) receptors. Docking is performed in two steps: (a) The charged terminal peptide residues are located by randomly distributing multiple copies of each in volumes of approximately 150 A at either end of the binding groove, and then minimizing the system energy using a modified multiple-copy search algorithm. This is followed by (b) construction of the intervening chain using the multiple-copy bond-scaling-relaxation loop closure algorithm. In both steps, the copies tend to cluster and the size of the resulting clusters is proportional to the basin of attraction of the corresponding energy well. We show that native MHC-bound peptides have broad minima and, consequently, that misfolded, low-energy peptide conformations can be eliminated by restricting consideration to groups of peptides which cluster into broad minima. The accuracy of the method is assessed by comparing the predictions with crystallographic data for three different MHC peptide systems, at various degrees of stringency: (a) the extent to which we can determine side chain function (anchor vs. T-cell epitopes); (b) the extent to which we can determine the peptide-receptor orientation; and (c) the accuracy with which we can predict atomic coordinates. We find the method correct on (a) for 19 of the 22 non-Gly positions; the failures appearing to be a consequence of omitting solvation. Predictions related to (b) are also very encouraging, with the overall orientation of the predicted peptides being very similar to the crystal conformation, when measured by the hydrogen bonding pattern between the two. The degree of success in predicting atomic coordinates varied considerably, however, from 1.4 A for the HLA-A2 peptide to 2.7 A for the Kb peptide. The inaccuracy of the latter appears to reflect an incomplete target function, most likely the ommission of solvation. The calculations thus define the current limits of accuracy in docking flexible peptides to Class I receptors and identify the methodological improvements that must be made for the next advance in accuracy.

Keywords: immunoinformatics

This paper describes a personal computer based tool to aid in decision making about whether a woman should receive adjuvant therapy for breast cancer. This tool can assist in engaging women with primary breast cancer in the discussion about: 1) her risk of breast cancer related mortality if she receives only local control measures, but no systemic adjuvant therapy, 2) how much receiving adjuvant therapy may reduce this risk, and 3) what the impact of receiving the adjuvant systemic therapy is in terms of survival. The tool utilizes life table analytical techniques to project outcomes after entry of patient age (used to calculate natural mortality rates), estimated risk of breast cancer related mortality (with a help tool allowing the physician to use estimates based on national database information), and estimate of the efficacy of adjuvant chemotherapy (with included tables of estimates based on the Early Breast Cancer Trialists' meta-analysis). Computer based tools can serve as valuable aids in patient and physician education, and the process of informed decision making.

Keywords: Aged; Breast Neoplasms, drug therapy/mortality; Chemotherapy, Adjuvant; Decision Making, Computer-Assisted; Education, Medical; Female; Humans; Life Tables; Middle Aged; Patient Participation; Prognosis; Software; Survival Rate

Keywords: immunoinformatics

To identify genes involved in cardiac arrhythmia, we investigated patients with long QT syndrome (LQT), an inherited disorder causing sudden death from a ventricular tachyarrythmia, torsade de pointes. We previously mapped LQT loci on chromosomes 11 (LQT1), 7 (LQT2), and 3 (LQT3). Here, linkage and physical mapping place LQT2 and a putative potassium channel gene, HERG, on chromosome 7q35-36. Single strand conformation polymorphism and DNA sequence analyses reveal HERG mutations in six LQT families, including two intragenic deletions, one splice-donor mutation, and three missense mutations. In one kindred, the mutation arose de novo. Northern blot analyses show that HERG is strongly expressed in the heart. These data indicate that HERG is LQT2 and suggest a likely cellular mechanism for torsade de pointes.

Keywords: herg

The support-vector network is a new learning machine for two-group classification problems. The machine conceptually implements the following idea: input vectors are non-linearly mapped to a very high-dimension feature space. In this feature space a linear decision surface is constructed. Special properties of the decision surface ensures high generalization ability of the learning machine. The idea behind the support-vector network was previously implemented for the restricted case where the training data can be separated without errors. We here extend this result to non-separable training data.

A knowledge of the fate of a drug, its disposition (absorption, distribution, metabolism, and excretion, known by the acronym ADME) and pharmacokinetics (the mathematical description of the rates of these processes and of concentration-time relationships), plays a central role throughout pharmaceutical research and development. These studies aid in the discovery and selection of new chemical entities, support safety assessment, and are critical in defining conditions for safe and effective use in patients. ADME studies provide the only basis for critical judgments from situations where the behavior of the drug is understood to those where it is unknown: this is most important in bridging from animal studies to the human situation. This presentation is intended to provide an introductory overview of the life cycle of a drug in the animal body and indicates the significance of such information for a full understanding of mechanisms of action and toxicity.

Keywords: chemogenomics

Describes a sequential universal data compression procedure for binary tree sources that performs the ?double mixture.? Using a context tree, this method weights in an efficient recursive way the coding distributions corresponding to all bounded memory tree sources, and achieves a desirable coding distribution for tree sources with an unknown model and unknown parameters. Computational and storage complexity of the proposed procedure are both linear in the source sequence length. The authors derive a natural upper bound on the cumulative redundancy of the method for individual sequences. The three terms in this bound can be identified as coding, parameter, and model redundancy, The bound holds for all source sequence lengths, not only for asymptotically large lengths. The analysis that leads to this bound is based on standard techniques and turns out to be extremely simple. The upper bound on the redundancy shows that the proposed context-tree weighting procedure is optimal in the sense that it achieves the Rissanen (1984) lower bound

An irreducible parameterization for a finite memory source is constructed in the form of a tree machine. A universal information source for the set of finite memory sources is constructed by a predictive modification of an earlier studied algorithm-Context. It is shown that this universal source incorporates any minimal data-generating tree machine in an asymptotically optimal manner in the following sense: the negative logarithm of the probability it assigns to any long typical sequence, generated by any tree machine, approaches that assigned by the tree machine at the best possible rate

Keywords: information-theory

The capacity of the channel induced by a given class of sources is well known to be an attainable lower bound on the redundancy of universal codes with respect to this class, both in the minimax sense and in the Bayesian (maximin) sense. We show that this capacity is essentially a lower bound also in a stronger sense, that is, for ?most? sources in the class. This result extends Rissanen's (1984, 1986) lower bound for parametric families. We demonstrate the applicability of this result in several examples, e.g., parametric families with growing dimensionality, piecewise-fixed sources, arbitrarily varying sources, and noisy samples of learnable functions. Finally, we discuss implications of our results to statistical inference

Keywords: information-theory

A general notion of universal consistency of nonparametric estimators is introduced that applies to regression estimation, conditional median estimation, curve fitting, pattern recognition, and learning concepts. General methods for proving consistency of estimators based on minimizing the empirical error are shown. In particular, distribution-free almost sure consistency of neural network estimates and generalized linear estimators is established

Keywords: information-theory

The article focuses on lower bound results on expected redundancy for universal coding of independent and identically distributed data on [0, 1] from parametric and nonparametric families. After reviewing existing lower bounds, we provide a new proof for minimax lower bounds on expected redundancy over nonparametric density classes. This new proof is based on the calculation of a mutual information quantity, or it utilizes the relationship between redundancy and Shannon capacity. It therefore unifies the minimax redundancy lower bound proofs in the parametric and nonparametric cases

Keywords: information-theory

Separation and identification of proteins by two-dimensional (2-D) electrophoresis can be used for protein-based gene expression analysis. In this report single protein spots, from polyvinylidene difluoride blots of micropreparative E. coli 2-D gels, were rapidly and economically identified by matching their amino acid composition, estimated pI and molecular weight against all E. coli entries in the SWISS-PROT database. Thirty proteins from an E. coli 2-D map were analyzed and identities assigned. Three of the proteins were unknown. By protein sequencing analysis, 20 of the 27 proteins were correctly identified. Importantly, correct identifications showed unambiguous "correct" score patterns. While incorrect protein identifications also showed distinctive score patterns, indicating that protein must be identified by other means. These techniques allow large-scale screening of the protein complement of simple organisms, or tissues in normal and disease states. The computer program described here is accessible via the World Wide Web at URL address (http:@expasy.hcuge.ch/).

Keywords: Amino Acids; Bacterial Proteins; Blood Proteins; Databases, Factual; Electrophoresis, Gel, Two-Dimensional; Escherichia coli; Humans; Microchemistry; Molecular Weight; Multienzyme Complexes; Proteins; Reproducibility of Results; Software; Time Factors

An HLA-A3-like supertype (minimally comprised of products from the HLA class I alleles A3, A11, A31, A*3301, and A*6801) has been defined on the basis of (a) structural similarities in the antigen-binding groove, (b) shared main anchor peptide-binding motifs, (c) the identification of peptides cross-reacting with most or all of these molecules, and (d) the definition of an A3-like supermotif that efficiently predicts highly cross-reactive peptides. Detailed secondary anchor maps for A3, A11, A31, A*3301, and A*6801 are also described. The biologic relevance of the A3-like supertype is indicated by the fact that high frequencies of the A3-like supertype alleles are conserved in all major ethnic groups. Because A3-like supertype alleles are found in most major HLA evolutionary lineages, possibly a reflection of common ancestry, the A3-like supermotif might in fact represent a primeval human HLA class I peptide-binding specificity. It is also possible that these phenomena might be related to optimal exploitation of the peptide specificity by human TAP molecules. The grouping of HLA alleles into supertypes on the basis of their overlapping peptide-binding repertoires represents an alternative to serologic or phylogenetic classification.

Keywords: Alleles; Amino Acid Sequence; Cell Line, Transformed; Cross Reactions; HLA Antigens; HLA-A3 Antigen; HLA-B Antigens; Haplotypes; Humans; Molecular Sequence Data; Peptide Fragments; Protein Binding; Structure-Activity Relationship

By taking into account the Fisher information and removing an inherent redundancy in earlier two-part codes, a sharper code length as the stochastic complexity and the associated universal process are derived for a class of parametric processes. The main condition required is that the maximum-likelihood estimates satisfy the central limit theorem. The same code length is also obtained from the so-called maximum-likelihood code

Keywords: information-theory

This report describes a computer program designed to assist health care professionals in making projections of the average benefit of systemic adjuvant therapy for individual breast cancer patients. It requires as input patient age (used to make projections of natural mortality), an estimate of breast cancer-related mortality at 5 years (used to make projections of breast cancer-specific mortality), and the proportional risk reduction for breast cancer mortality expected for the adjuvant therapy (with included tables from the Early Breast Cancer Trialist's 1992 meta-analysis). The program uses life table analytical techniques to make projections of outcome in three scenarios: that the breast cancer never occurred, that the breast cancer patient received definitive local therapy but no adjuvant systemic therapy, and that the patient received adjuvant therapy. The outcome projections are given for total, natural (non-breast cancer-related), and breast cancer-related mortality at several time points and also of total remaining life expectancy. These estimates are currently widely made by clinicians by nonnumerical techniques. Computer-based tools can serve as valuable aids in physician and patient education and in the process of informed decision making.

Keywords: Adult; Age Factors; Aged; Aged, 80 and over; Breast Neoplasms, drug therapy/mortality/surgery; Chemotherapy, Adjuvant; Decision Making; Female; Humans; Life Expectancy; Life Tables; Middle Aged; Proportional Hazards Models; Risk Factors; SEER Program; Software; Survival Rate; Treatment Outcome

We present an algorithm for placing molecular fragments into the active site of a receptor. A molecular fragment is defined as a connected part of a molecule containing only complete ring systems. The algorithm is part of a docking tool, called FLEXX, which is currently under development at GMD. The overall goal is to provide means of automatically computing low-energy conformations of the ligand within the active site, with an accuracy approaching the limitations of experimental methods for resolving molecular structures and within a run time that allows for docking large sets of ligands. The methods by which we plan to achieve this goal are the explicit exploitation of molecular flexibility of the ligand and the incorporation of physicochemical properties of the molecules. The algorithm for fragment placement, which is the topic of this paper, is based on pattern recognition techniques and is able to predict a small set of possible positions of a molecular fragment with low flexibility within seconds on a workstation. In most cases, a placement with rms deviation below 1.0 A with respect to the X-ray structure is found among the 10 highest ranking solutions, assuming that the receptor is given in the bound conformation.

Keywords: Algorithms; Binding Sites; Databases, Factual; Ligands; Models, Chemical; Peptide Fragments, chemistry; Proteins, chemistry; Software

We present an automatic method for docking organic ligands into protein binding sites. The method can be used in the design process of specific protein ligands. It combines an appropriate model of the physico-chemical properties of the docked molecules with efficient methods for sampling the conformational space of the ligand. If the ligand is flexible, it can adopt a large variety of different conformations. Each such minimum in conformational space presents a potential candidate for the conformation of the ligand in the complexed state. Our docking method samples the conformation space of the ligand on the basis of a discrete model and uses a tree-search technique for placing the ligand incrementally into the active site. For placing the first fragment of the ligand into the protein, we use hashing techniques adapted from computer vision. The incremental construction algorithm is based on a greedy strategy combined with efficient methods for overlap detection and for the search of new interactions. We present results on 19 complexes of which the binding geometry has been crystallographically determined. All considered ligands are docked in at most three minutes on a current workstation. The experimentally observed binding mode of the ligand is reproduced with 0.5 to 1.2 A rms deviation. It is almost always found among the highest-ranking conformations computed.

Keywords: Aldehyde Reductase, Algorithms, Amiloride, Aminoimidazole Carboxamide, Animals, Arabinose, Automation, Binding Sites, Carbonic Anhydrases, Computational Biology, Computer Simulation, Concanavalin A, Crystallography, Databases, Drug Design, Drug Evaluation, Enzyme Inhibitors, Factual, Folic Acid, Folic Acid Antagonists, Fructose-Bisphosphatase, Humans, Internet, Ligands, Methotrexate, Models, Molecular, Non-U.S. Gov't, Pancreatic Elastase, Pentamidine, Pliability, Point Mutation, Preclinical, Protein Binding, Protein Conformation, Proteins, Reproducibility of Results, Research Support, Ribonucleosides, Software, Tetrahydrofolate Dehydrogenase, Thermolysin, Time Factors, Trypsin, X-Ray, 8780787

The current interest in combinatorial chemistry for lead generation has necessitated the development of methods for design and evaluation of the diversity of the resultant compound libraries. Such methods also have application in selecting diverse sets of compounds for general screening from corporate databases and in the analysis of large sets of structures to identify common patterns. In this paper we describe a novel methodology for calculating diversity and identifying common features based on the three-point pharmacophores expressed by a compound.1 The method has been implemented within the environment of the Chem-X molecular modeling package (ChemDBS-3D), using a systematic analysis of 3D distance space with three point combinations of six pharmacophoric groups. The strategy used to define the pharmacophores is discussed, including an in-house developed atom type parameterization. The method is compared with the related approach being developed into the ChemDiverse module of Chem-X. Results from an analysis of a large corporate database and examples of combinatorial library profiling with both methods are presented. The use of 3D pharmacophores for assessing diversity, and the application of such methods to combinatorial library design, is discussed.

Keywords: chemoinformatics

The minimum complexity regression estimation framework (Barron, 1991; Barron and Cover, 1991 and Rissanen, 1989) is a general data-driven methodology for estimating a regression function from a given list of parametric models using independent and identically distributed (i.i.d.) observations. We extend Barron's regression estimation framework to m-dependent observations and to strongly mixing observations. In particular, we propose abstract minimum complexity regression estimators for dependent observations, which may be adapted to a particular list of parametric models, and establish upper bounds on the statistical risks of the proposed estimators in terms of certain deterministic indices of resolvability. Assuming that the regression function satisfies a certain Fourier-transform-type representation, we examine minimum complexity regression estimators adapted to a list of parametric models based on neural networks and by using the upper bounds for the abstract estimators, we establish rates of convergence for the statistical risks of these estimators. Also, as a key tool, we extend the classical Bernstein inequality from i.i.d. random variables to m-dependent processes and to strongly mixing processes

Keywords: information-theory

In pattern recognition or, as it has also been called, concept learning, the value of a 0,1-valued random variable Y is to be predicted based upon observing an Rd-valued random variable X. We apply the method of complexity regularization to learn concepts from large concept classes. The method is shown to automatically find a good balance between the approximation error and the estimation error. In particular, the error probability of the obtained classifier is shown to decrease as O(?(logn/n)) to the achievable optimum, for large nonparametric classes of distributions, as the sample size n grows. We also show that if the Bayes error probability is zero and the Bayes rule is in a known family of decision rules, the error probability is O(logn/n) for many large families, possibly with infinite VC dimension

Keywords: information-theory

A database system and computer programs for storage and retrieval of information about guanine nucleotide-binding protein (G protein) -coupled receptor mutants and associated biological effects have been developed. Mutation data on the receptors were collected from the literature and a database of mutants and effects of mutations was developed. The G protein-coupled receptor, family A, point mutation database (GRAP) provides detailed information on ligand-binding and signal transduction properties of more than 2130 receptor mutants. The amino acid sequences of receptors for which mutation experiments have been reported were aligned, and from this alignment mutation data may be retrieved. Alternatively, a search form allowing detailed specification of which mutants to retrieve may be used, for example, to search for specific amino acid substitutions, substitutions in specific protein domains or reported biological effects. Furthermore, ligand and bibliographic oriented queries may be performed. GRAP is available on the Internet (URL: http://www-grap.fagmed.uit.no/GRAP/+ +homepage.html) using the World-Wide Web system.

Keywords: Amino Acid Sequence; Computer Communication Networks; Computers; GTP-Binding Proteins; Information Systems; Molecular Sequence Data; Mutagenesis, Site-Directed; Mutation; Receptors, Cell Surface; Sequence Alignment

We present a general approach to forming structure-activity relationships (SARs). This approach is based on representing chemical structure by atoms and their bond connectivities in combination with the inductive logic programming (ILP) algorithm PROGOL. Existing SAR methods describe chemical structure by using attributes which are general properties of an object. It is not possible to map chemical structure directly to attribute-based descriptions, as such descriptions have no internal organization. A more natural and general way to describe chemical structure is to use a relational description, where the internal construction of the description maps that of the object described. Our atom and bond connectivities representation is a relational description. ILP algorithms can form SARs with relational descriptions. We have tested the relational approach by investigating the SARs of 230 aromatic and heteroaromatic nitro compounds. These compounds had been split previously into two subsets, 188 compounds that were amenable to regression and 42 that were not. For the 188 compounds, a SAR was found that was as accurate as the best statistical or neural network-generated SARs. The PROGOL SAR has the advantages that it did not need the use of any indicator variables handcrafted by an expert, and the generated rules were easily comprehensible. For the 42 compounds, PROGOL formed a SAR that was significantly (P < 0.025) more accurate than linear regression, quadratic regression, and back-propagation. This SAR is based on an automatically generated structural alert for mutagenicity.

VMD is a molecular graphics program designed for the display and analysis of molecular assemblies, in particular biopolymers such as proteins and nucleic acids. VMD can simultaneously display any number of structures using a wide variety of rendering styles and coloring methods. Molecules are displayed as one or more "representations," in which each representation embodies a particular rendering method and coloring scheme for a selected subset of atoms. The atoms displayed in each representation are chosen using an extensive atom selection syntax, which includes Boolean operators and regular expressions. VMD provides a complete graphical user interface for program control, as well as a text interface using the Tcl embeddable parser to allow for complex scripts with variable substitution, control loops, and function calls. Full session logging is supported, which produces a VMD command script for later playback. High-resolution raster images of displayed molecules may be produced by generating input scripts for use by a number of photorealistic image-rendering applications. VMD has also been expressly designed with the ability to animate molecular dynamics (MD) simulation trajectories, imported either from files or from a direct connection to a running MD simulation. VMD is the visualization component of MDScope, a set of tools for interactive problem solving in structural biology, which also includes the parallel MD program NAMD, and the MDCOMM software used to connect the visualization and simulation programs. VMD is written in C++, using an object-oriented design; the program, including source code and extensive documentation, is freely available via anonymous ftp and through the World Wide Web.

In an earlier paper, we proved a strong version of the redundancy-capacity converse theorem of universal coding, stating that for ?most? sources in a given class, the universal coding redundancy is essentially lower-bounded by the capacity of the channel induced by this class. Since this result holds for general classes of sources, it extends Rissanen's (1986) strong converse theorem for parametric families. While our earlier result has established strong optimality only for mixture codes weighted by the capacity-achieving prior, our first result herein extends this finding to a general prior. For some cases our technique also leads to a simplified proof of the above mentioned strong converse theorem. The major interest in this paper, however, is in extending the theory of universal coding to hierarchical structures of classes, where each class may have a different capacity. In this setting, one wishes to incur redundancy essentially as small as that corresponding to the active class, and not the union of classes. Our main result is that the redundancy of a code based on a two-stage mixture (first, within each class, and then over the classes), is no worse than that of any other code for ?most? sources of ?most? classes. If, in addition, the classes can be efficiently distinguished by a certain decision rule, then the best attainable redundancy is given explicitly by the capacity of the active class plus the normalized negative logarithm of the prior probability assigned to this class. These results suggest some interesting guidelines as for the choice of the prior. We also discuss some examples with a natural hierarchical partition into classes

Keywords: information-theory source-coding

We present a sequential investment algorithm, the ?-weighted universal portfolio with side information, which achieves, to first order in the exponent, the same wealth as the best side-information dependent investment strategy (the best state-constant rebalanced portfolio) determined in hindsight from observed market and side-information outcomes. This is an individual sequence result which shows the difference between the exponential growth wealth of the best state-constant rebalanced portfolio and the universal portfolio with side information is uniformly less than (d/(2n))log (n+1)+(k/n)log 2 for every stock market and side-information sequence and for all time n. Here d=k(m-1) is the number of degrees of freedom in the state-constant rebalanced portfolio with k states of side information and m stocks. The proof of this result establishes a close connection between universal investment and universal data compression

Keywords: information-theory

The molecular mechanisms involved in transgene-induced gene silencing ('quelling') in Neurospora crassa were investigated using the carotenoid biosynthetic gene albino-1 (al-1) as a visual marker. Deletion derivatives of the al-1 gene showed that a transgene must contain at least approximately 132 bp of sequences homologous to the transcribed region of the native gene in order to induce quelling. Transgenes containing only al-1 promoter sequences do not cause quelling. Specific sequences are not required for gene silencing, as different regions of the al-1 gene produced quelling. A mutant defective in cytosine methylation (dim-2) exhibited normal frequencies and degrees of silencing, indicating that cytosine methylation is not responsible for quelling, despite the fact that methylation of transgene sequences frequently is correlated with silencing. Silencing was shown to be a dominant trait, operative in heterokaryotic strains containing a mixture of transgenic and non-transgenic nuclei. This result indicates that a diffusable, trans-acting molecule is involved in quelling. A transgene-derived, sense RNA was detected in quelled strains and was found to be absent in their revertants. These data are consistent with a model in which an RNA-DNA or RNA-RNA interaction is involved in transgene-induced gene silencing in Neurospora.

Keywords: sirna

Keywords: chemoinformatics

Keywords: PUlearning

Keywords: csbcbook, csbcbook-ch3

Post-transcriptional gene silencing in transgenic plants is the manifestation of a mechanism that suppresses RNA accumulation in a sequence-specific manner. The target RNA species may be the products of transgenes, endogenous plant genes or viral RNAs. For an RNA to be a target it is necessary only that it has sequence homology to the sense RNA product of the transgene. There are three current hypotheses to account for the mechanism of post transcriptional gene silencing. These models all require production of an antisense RNA of the RNA targets to account for the specificity of the mechanism. There could be either direct transcription of the antisense RNA from the transgene, antisense RNA produced in response to over expression of the transgene or antisense RNA produced in response to the production of an aberrant sense RNA product of the transgene. To determine which of these models is correct it will be necessary to find out whether transgene methylation, which is frequently associated with the potential of transgenes to confer post-transcriptional gene silencing, is a cause or a consequence of the process.

Electronic properties located on the atoms of a molecule such as partial atomic charges as well as electronegativity and polarizability values are encoded by an autocorrelation vector accounting for the constitution of a molecule. This encoding procedure is able to distinguish between compounds being dopamine agonists and those being benzodiazepine receptor agonists even after projection into a two-dimensional self-organizing network. The two types of compounds can still be distinguished if they are buried in a dataset of 8323 compounds of a chemical supplier catalog comprising a wide structural variety. The maps obtained by this sequence of events, calculation of empirical physicochemical effects, encoding in a topological autocorrelation vector, and projection by a self-organizing neural network, can thus be used for searching for structural similarity, and, in particular, for finding new lead structures with biological activity.

Keywords: Animals, Chemical, Chemistry, Databases, Dopamine Agonists, Drug Design, Electrochemistry, Factual, GABA-A, Logistic Models, Models, Molecular Structure, Neural Networks (Computer), Non-U.S. Gov't, Phenols, Physical, Receptors, Research Support, Structure-Activity Relationship, Tetrahymena pyriformis, 8941996

A graduated assignment algorithm for graph matching is presented which is fast and accurate even in the presence of high noise. By combining graduated nonconvexity, two-way (assignment) constraints, and sparsity, large improvements in accuracy and speed are achieved. Its low order computational complexity [O(lm), where l and m are the number of links in the two graphs] and robustness in the presence of noise offer advantages over traditional combinatorial approaches. The algorithm, not restricted to any special class of graph, is applied to subgraph isomorphism, weighted graph matching, and attributed relational graph matching. To illustrate the performance of the algorithm, attributed relational graphs derived from objects are matched. Then, results from twenty-five thousand experiments conducted on 100 node random graphs of varying types (graphs with only zero-one links, weighted graphs, and graphs with node attributes and multiple link types) are reported. No comparable results have been reported by any other graph matching algorithm before in the research literature. Twenty-five hundred control experiments are conducted using a relaxation labeling algorithm and large improvements in accuracy are demonstrated.

Let Xn=(X1,...,Xn) be a memoryless source with unknown distribution on a finite alphabet of size k. We identify the asymptotic minimax coding redundancy for this class of sources, and provide a sequence of asymptotically minimax codes. Equivalently, we determine the limiting behavior of the minimax relative entropy minQXn maxpXn D(PXn∥QXn), where the maximum is over all independent and identically distributed (i.i.d.) source distributions and the minimum is over all joint distributions. We show in this paper that the minimax redundancy minus ((k-1)/2) log(n/(2?e)) converges to log??(det I(&thetas;))d&thetas;=log (?(1/2)k/?(k/2)), where I(&thetas;) is the Fisher information and the integral is over the whole probability simplex. The Bayes strategy using Jeffreys' prior is shown to be asymptotically maximin but not asymptotically minimax in our setting. The boundary risk using Jeffreys' prior is higher than that of interior points. We provide a sequence of modifications of Jeffreys' prior that put some prior mass near the boundaries of the probability simplex to pull down that risk to the asymptotic minimax level in the limit

Keywords: information-theory

I propose a new method for variable selection and shrinkage in Cox's proportional hazards model. My proposal minimizes the log partial likelihood subject to the sum of the absolute values of the parameters being bounded by a constant. Because of the nature of this constraint, it shrinks coefficients and produces some coefficients that are exactly zero. As a result it reduces the estimation variance while providing an interpretable final model. The method is a variation of the 'lasso' proposal of Tibshirani, designed for the linear regression context. Simulations indicate that the lasso can be more accurate than stepwise selection in this setting.

Databases of multiple sequence alignments are a valuable aid to protein sequence classification and analysis. One of the main challenges when constructing such a database is to simultaneously satisfy the conflicting demands of completeness on the one hand and quality of alignment and domain definitions on the other. The latter properties are best dealt with by manual approaches, whereas completeness in practice is only amenable to automatic methods. Herein we present a database based on hidden Markov model profiles (HMMs), which combines high quality and completeness. Our database, Pfam, consists of parts A and B. Pfam-A is curated and contains well-characterized protein domain families with high quality alignments, which are maintained by using manually checked seed alignments and HMMs to find and align all members. Pfam-B contains sequence families that were generated automatically by applying the Domainer algorithm to cluster and align the remaining protein sequences after removal of Pfam-A domains. By using Pfam, a large number of previously unannotated proteins from the Caenorhabditis elegans genome project were classified. We have also identified many novel family memberships in known proteins, including new kazal, Fibronectin type III, and response regulator receiver domains. Pfam-A families have permanent accession numbers and form a library of HMMs available for searching and automatic annotation of new protein sequences.

Keywords: csbcbook, csbcbook-ch2

The conditional distribution of the next outcome given the infinite past of a stationary process can be inferred from finite but growing segments of the past. Several schemes are known for constructing pointwise consistent estimates, but they all demand prohibitive amounts of input data. We consider real-valued time series and construct conditional distribution estimates that make much more efficient use of the input data. The estimates are consistent in a weak sense, and the question whether they are pointwise-consistent is still open. For finite-alphabet processes one may rely on a universal data compression scheme like the Lempel-Ziv (1978) algorithm to construct conditional probability mass function estimates that are consistent in expected information divergence. Consistency in this strong sense cannot be attained in a universal sense for all stationary processes with values in an infinite alphabet, but weak consistency can. Some applications of the estimates to on-line forecasting, regression, and classification are discussed

Keywords: information-theory

Here we demonstrate the use of a mammalian two-hybrid system to study protein-protein interactions. Like the yeast two-hybrid system, this is a genetic, in vivo assay based on the reconstitution of the function of a transcriptional activator. In this system, one protein of interest is expressed as a fusion to the Gal4 DNA-binding domain and another protein is expressed as a fusion to the activation domain of the VP16 protein of the herpes simplex virus. The vectors that express these fusion proteins are cotransfected with a reporter chloramphenicol acetyltransferase (CAT) vector into a mammalian cell line. The reporter plasmid contains a cat gene under the control of five consensus Gal4 binding sites. If the two fusion proteins interact, there will be a significant increase in expression of the cat reporter gene. Previously, it was reported that mouse p53 antitumor protein and simian virus 40 large T antigen interact in a yeast two-hybrid system. Using a mammalian two-hybrid system, we were able to independently confirm this interaction. The mammalian two-hybrid system can be used as a complementary approach to verify protein-protein interactions detected by a yeast two-hybrid system screening. In addition, the mammalian two-hybrid system has two main advantages: (i) Assay results can be obtained within 48 h of transfection, and (ii) protein interactions in mammalian cells may better mimic actual in vivo interactions.

Keywords: Antigens, Polyomavirus Transforming; Binding Sites; Chloramphenicol O-Acetyltransferase; DNA; DNA-Binding Proteins; Fungal Proteins; Genes, Reporter; Genetic Vectors; Hela Cells; Herpes Simplex Virus Protein Vmw65; Humans; Promoter Regions, Genetic; Recombinant Fusion Proteins; Saccharomyces cerevisiae Proteins; Simian virus 40; Transcription Factors; Transfection; Tumor Suppressor Protein p53

Keywords: lasso, ordinal, regression

Prediction of small molecule binding modes to macromolecules of known three-dimensional structure is a problem of paramount importance in rational drug design (the "docking" problem). We report the development and validation of the program GOLD (Genetic Optimisation for Ligand Docking). GOLD is an automated ligand docking program that uses a genetic algorithm to explore the full range of ligand conformational flexibility with partial flexibility of the protein, and satisfies the fundamental requirement that the ligand must displace loosely bound water on binding. Numerous enhancements and modifications have been applied to the original technique resulting in a substantial increase in the reliability and the applicability of the algorithm. The advanced algorithm has been tested on a dataset of 100 complexes extracted from the Brookhaven Protein DataBank. When used to dock the ligand back into the binding site, GOLD achieved a 71% success rate in identifying the experimental binding mode.

Keywords: Algorithms, Binding Sites, Computer Simulation, Crystallography, Genetic, Humans, Ligands, Models, Molecular, NADP, Protein Binding, Protein Conformation, Proteins, Tetrahydrofolate Dehydrogenase, X-Ray, 9126849

This article describes a program for pharmacophore mapping, called MPHIL (Mapping Pharmacophores in Ligands). Given as input a set of molecules that exhibit some common biological activity, MPHIL identifies the smallest 3D pattern of pharmacophore points that has at least m (a user-defined parameter) points in common with each of the input molecules. The program thus differs from existing programs for pharmacophore mapping in that it does not require all of the molecules to share exactly the same pattern of points, although it will find such a common pattern if it does, indeed, exist. MPHIL uses a genetic algorithm (GA) approach in which an initial, and very rapid, GA is used to suggest possible combinations of points that are then processed by the second GA to yield the final 3D pattern.

Keywords: chemoinformatics

The amount of fixed side information required for lossless data compression is discussed. Nonasymptotic coding and converse theorems are derived for data-compression algorithms with fixed statistical side information (?training sequence?) that is not large enough so as to yield the ultimate compression, namely, the entropy of the source

Keywords: information-theory

Suppose nature picks a probability measure P&thetas; on a complete separable metric space X at random from a measurable set P ?=P&thetas;:&thetas;??. Then, without knowing &thetas;, a statistician picks a measure Q on S. Finally, the statistician suffers a loss D(P0∥Q), the relative entropy between P&thetas; and Q. We show that the minimax and maximin values of this game are always equal, and there is always a minimax strategy in the closure of the set of all Bayes strategies. This generalizes previous results of Gallager(1979), and Davisson and Leon-Garcia (1980)

Peptides that bind to major histocompatibility complex products (MHC) are known to exhibit certain sequence motifs which, though common, are neither necessary nor sufficient for binding: MHCs bind certain peptides that do not have the characteristic motifs and only about 30% of the peptides having the required motif, bind. In order to develop and test more accurate methods we measured the binding affinity of 463 nonamer peptides to HLA-A2.1. We describe two methods for predicting whether a given peptide will bind to an MHC and apply them to these peptides. One method is based on simulating a neural network and another, called the polynomial method, is based on statistical parameter estimation assuming independent binding of the side-chains of residues. We compare these methods with each other and with standard motif-based methods. The two methods are complementary, and both are superior to sequence motifs. The neural net is superior to simple motif searches in eliminating false positives. Its behavior can be coarsely tuned to the strength of binding desired and it is extendable in a straightforward fashion to other alleles. The polynomial method, on the other hand, has high sensitivity and is a superior method for eliminating false negatives. We discuss the validity of the independent binding assumption in such predictions.

Keywords: Artificial Intelligence; Computing Methodologies; HLA-A2 Antigen; Neural Networks (Computer); Oligopeptides; Protein Binding; Reproducibility of Results

Keywords: chemoinformatics

Cells use complex networks of interacting molecular components to transfer and process information. These "computational devices of living cells" are responsible for many important cellular processes, including cell-cycle regulation and signal transduction. Here we address the issue of the sensitivity of the networks to variations in their biochemical parameters. We propose a mechanism for robust adaptation in simple signal transduction networks. We show that this mechanism applies in particular to bacterial chemotaxis. This is demonstrated within a quantitative model which explains, in a unified way, many aspects of chemotaxis, including proper responses to chemical gradients. The adaptation property is a consequence of the network's connectivity and does not require the 'fine-tuning' of parameters. We argue that the key properties of biochemical networks should be robust in order to ensure their proper functioning.

A method for the prediction of the interactions within complex reaction networks from experimentally measured time series of the concentration of the species composing the system has been tested experimentally on the first few steps of the glycolytic pathway. The reconstituted reaction system, containing eight enzymes and 14 metabolic intermediates, was kept away from equilibrium in a continuous-flow, stirred-tank reactor. Input concentrations of adenosine monophosphate and citrate were externally varied over time, and their concentrations in the reactor and the response of eight other species were measured. Multidimensional scaling analysis and heuristic algorithms applied to two-species time-lagged correlation functions derived from the time series yielded a diagram from which the interactions among all of the species could be deduced. The diagram predicts essential features of the known reaction network in regard to chemical reactions and interactions among the measured species. The approach is applicable to many complex reaction systems.

Keywords: reconstruction, kinetic, metabolism, analysis, multidimensional scaling, Correlation Metric Construction

Improved thermodynamic parameters for prediction of RNA duplex formation are derived from optical melting studies of 90 oligoribonucleotide duplexes containing only Watson-Crick base pairs. To test end or base composition effects, new sets of duplexes are included that have identical nearest neighbors, but different base compositions and therefore different ends. Duplexes with terminal GC pairs are more stable than duplexes with the same nearest neighbors but terminal AU pairs. Penalizing terminal AU base pairs by 0.45 kcal/mol relative to terminal GC base pairs significantly improves predictions of DeltaG degrees37 from a nearest-neighbor model. A physical model is suggested in which the differential treatment of AU and GC ends accounts for the dependence of the total number of Watson-Crick hydrogen bonds on the base composition of a duplex. On average, the new parameters predict DeltaG degrees37, DeltaH degrees, DeltaS degrees, and TM within 3.2%, 6.0%, 6.8%, and 1.3 degreesC, respectively. These predictions are within the limit of the model, based on experimental results for duplexes predicted to have identical thermodynamic parameters.

A large fraction of HLA class I, and possibly class II, molecules can be classified into relatively few supertypes, characterized by overlapping peptide-binding repertoires and consensus B- and F-pocket structures. Cross-binding peptides are frequently recognized by specific T cells in the course of natural disease processes and in the context of multiple HLA molecules, validating the concept of HLA supertypes at the functional level.

Keywords: Animals; Communicable Diseases; Evolution, Molecular; HLA Antigens; HLA-A2 Antigen; HLA-A3 Antigen; Humans; Neoplasms; Polymorphism, Genetic

The theory of artificial neural networks is briefly reviewed focusing on supervised and unsupervised techniques which have great impact on current chemical applications. An introduction to molecular descriptors and representation schemes is given. In addition, worked examples of recent advances in this field are highlighted and pioneering publications are discussed. Applications of several types of artificial neural networks to compound classification, modelling of structure-activity relationships, biological target identification, and feature extraction from biopolymers are presented and compared to other techniques. Advantages and limitations of neural networks for computer-aided molecular design and sequence analysis are discussed.

Keywords: Algorithms, Amino Acid Sequence, Amino Acids, Animals, Artificial Intelligence, Automated, Bacterial, Bacterial Proteins, Bicuculline, Binding Sites, Biological, Biological Availability, Blood Proteins, Blood-Brain Barrier, Cation Transport Proteins, Cats, Cell Membrane Permeability, Chemical, Chemistry, Cluster Analysis, Combinatorial Chemistry Techniques, Comparative Study, Computational Biology, Computer Simulation, Computer Systems, Computer-Aided Design, Computer-Assisted, Computing Methodologies, DNA-Binding Proteins, Databases, Dogs, Drug Design, Electric Stimulation, Electromyography, Enzyme Inhibitors, Ether-A-Go-Go Potassium Channels, Excitatory Amino Acid Antagonists, Factual, False Positive Reactions, Forecasting, Forelimb, GABA Antagonists, Gene Expression Profiling, Genome, Glutamic Acid, Humans, Hydrogen Bonding, Image Enhancement, Image Interpretation, Image Processing, Information Storage and Retrieval, Iontophoresis, Kynurenic Acid, Least-Squares Analysis, Linear Models, Liver, Markov Chains, Metabolic Clearance Rate, Metalloendopeptidases, Microelectrodes, Models, Molecular, Molecular Conformation, Molecular Sequence Data, Molecular Structure, Motor Cortex, Movement, Multivariate Analysis, Nerve Net, Neural Networks (Computer), Neuropeptides, Non-U.S. Gov't, Nonlinear Dynamics, Pattern Recognition, Pharmaceutical, Pharmaceutical Preparations, Pharmacokinetics, Phylogeny, Potassium Channels, Predictive Value of Tests, Protein Interaction Mapping, Protein Sorting Signals, Protein Structure, Proteins, Rats, Reproducibility of Results, Research Support, Sensitivity and Specificity, Sequence Alignment, Sequence Analysis, Shoulder, Signal Processing, Software, Statistical, Stereotaxic Techniques, Structure-Activity Relationship, Terminology, Tertiary, Trans-Activators, Voltage-Gated, Zinc, 9830312

We derive a new general representation for a function as a linear combination of local correlation kernels at optimal sparse locations (and scales) and characterize its relation to principal component analysis, regularization, sparsity principles, and support vector machines.

Keywords: Algorithms, Automated, Biometry, Computers, DNA, Databases, Factual, Fungal, Fungal Proteins, GTP-Binding Proteins, Gene Expression, Genes, Learning, Markov Chains, Models, Neural Networks (Computer), Neurological, Non-P.H.S., Non-U.S. Gov't, Nucleic Acid Hybridization, Open Reading Frames, P.H.S., Pattern Recognition, Protein, Protein Structure, Proteins, Reproducibility of Results, Research Support, Saccharomyces cerevisiae, Sequence Alignment, Sequence Analysis, Software, Statistical, Tertiary, U.S. Gov't, 9698352

Gene dosage variations occur in many diseases. In cancer, deletions and copy number increases contribute to alterations in the expression of tumour-suppressor genes and oncogenes, respectively. Developmental abnormalities, such as Down, Prader Willi, Angelman and Cri du Chat syndromes, result from gain or loss of one copy of a chromosome or chromosomal region. Thus, detection and mapping of copy number abnormalities provide an approach for associating aberrations with disease phenotype and for localizing critical genes. Comparative genomic hybridization (CGH) was developed for genome-wide analysis of DNA sequence copy number in a single experiment. In CGH, differentially labelled total genomic DNA from a 'test' and a 'reference' cell population are cohybridized to normal metaphase chromosomes, using blocking DNA to suppress signals from repetitive sequences. The resulting ratio of the fluorescence intensities at a location on the 'cytogenetic map', provided by the chromosomes, is approximately proportional to the ratio of the copy numbers of the corresponding DNA sequences in the test and reference genomes. CGH has been broadly applied to human and mouse malignancies. The use of metaphase chromosomes, however, limits detection of events involving small regions (of less than 20 Mb) of the genome, resolution of closely spaced aberrations and linking ratio changes to genomic/genetic markers. Therefore, more laborious locus-by-locus techniques have been required for higher resolution studies. Hybridization to an array of mapped sequences instead of metaphase chromosomes could overcome the limitations of conventional CGH (ref. 6) if adequate performance could be achieved. Copy number would be related to the test/reference fluorescence ratio on the array targets, and genomic resolution could be determined by the map distance between the targets, or by the length of the cloned DNA segments. We describe here our implementation of array CGH. We demonstrate its ability to measure copy number with high precision in the human genome, and to analyse clinical specimens by obtaining new information on chromosome 20 aberrations in breast cancer.

Keywords: cgh, csbcbook

Keywords: csbcbook

Keywords: biosvm microarray

We consider the problem of one-step-ahead prediction of a real-valued, stationary, strongly mixing random process (Xi)i=-??. The best mean-square predictor of X0 is its conditional mean given the entire infinite past (Xi)i=-?-1. Given a sequence of observations X1, X2, XN, we propose estimators for the conditional mean based on sequences of parametric models of increasing memory and of increasing dimension, for example, neural networks and Legendre polynomials. The proposed estimators select both the model memory and the model dimension, in a data-driven fashion, by minimizing certain complexity regularized least squares criteria. When the underlying predictor function has a finite memory, we establish that the proposed estimators are memory-universal: the proposed estimators, which do not know the true memory, deliver the same statistical performance (rates of integrated mean-squared error) as that delivered by estimators that know the true memory. Furthermore, when the underlying predictor function does not have a finite memory, we establish that the estimator based on Legendre polynomials is consistent

Keywords: information-theory

Computational methods were used to predict the sequences of peptides that bind to the MHC class I molecule, K(b). The rules for predicting binding sequences, which are limited, are based on preferences for certain amino acids in certain positions of the peptide. It is apparent though, that binding can be influenced by the amino acids in all of the positions of the peptide. An artificial neural network (ANN) has the ability to simultaneously analyze the influence of all of the amino acids of the peptide and thus may improve binding predictions. ANNs were compared to statistically analyzed peptides for their abilities to predict the sequences of K(b) binding peptides. ANN systems were trained on a library of binding and nonbinding peptide sequences from a phage display library. Statistical and ANN methods identified strong binding peptides with preferred amino acids. ANNs detected more subtle binding preferences, enabling them to predict medium binding peptides. The ability to predict class I MHC molecule binding peptides is useful for immunolological therapies involving cytotoxic-T cells.

Keywords: immunoinformatics

This paper consists of an overview on universal prediction from an information-theoretic perspective. Special attention is given to the notion of probability assignment under the self-information loss function, which is directly related to the theory of universal data compression. Both the probabilistic setting and the deterministic setting of the universal prediction problem are described with emphasis on the analogy and the differences between results in the two settings

The binding of a major histocompatibility complex (MHC) molecule to a peptide originating in an antigen is essential to recognizing antigens in immune systems, and it has proved to be important to use computers to predict the peptides that will bind to an MHC molecule. The purpose of this paper is twofold: First, we propose to apply supervised learning of hidden Markov models (HMMs) to this problem, which can surpass existing methods for the problem of predicting MHC-binding peptides. Second, we generate peptides that have high probabilities to bind to a certain MHC molecule, based on our proposed method using peptides binding to MHC molecules as a set of training data. From our experiments, in a type of cross-validation test, the discrimination accuracy of our supervised learning method is usually approximately 2-15% better than those of other methods, including backpropagation neural networks, which have been regarded as the most effective approach to this problem. Furthermore, using an HMM trained for HLA-A2, we present new peptide sequences that are provided with high binding probabilities by the HMM and that are thus expected to bind to HLA-A2 proteins. Peptide sequences not shown in this paper but with rather high binding probabilities can be obtained from the author.

Keywords: immunoinformatics

Given the immanent gene expression mapping covering whole genomes during development, health and disease, we seek computational methods to maximize functional inference from such large data sets. Is it possible, in principle, to completely infer a complex regulatory network architecture from input/output patterns of its variables? We investigated this possibility using binary models of genetic networks. Trajectories, or state transition tables of Boolean nets, resemble time series of gene expression. By systematically analyzing the mutual information between input states and output states, one is able to infer the sets of input elements controlling each element or gene in the network. This process is unequivocal and exact for complete state transition tables. We implemented this REVerse Engineering ALgorithm (REVEAL) in a C program, and found the problem to be tractable within the conditions tested so far. For n = 50 (elements) and k = 3 (inputs per element), the analysis of incomplete state transition tables (100 state transition pairs out of a possible 10(15)) reliably produced the original rule and wiring sets. While this study is limited to synchronous Boolean networks, the algorithm is generalizable to include multi-state models, essentially allowing direct application to realistic biological data sets. The ability to adequately solve the inverse problem may enable in-depth analysis of complex dynamic systems in biology and other fields.

Keywords: information-theory source-coding

Many genes and signalling pathways controlling cell proliferation, death and differentiation, as well as genomic integrity, are involved in cancer development. New techniques, such as serial analysis of gene expression and cDNA microarrays, have enabled measurement of the expression of thousands of genes in a single experiment, revealing many new, potentially important cancer genes. These genome screening tools can comprehensively survey one tumor at a time; however, analysis of hundreds of specimens from patients in different stages of disease is needed to establish the diagnostic, prognostic and therapeutic importance of each of the emerging cancer gene candidates. Here we have developed an array-based high-throughput technique that facilitates gene expression and copy number surveys of very large numbers of tumors. As many as 1000 cylindrical tissue biopsies from individual tumors can be distributed in a single tumor tissue microarray. Sections of the microarray provide targets for parallel in situ detection of DNA, RNA and protein targets in each specimen on the array, and consecutive sections allow the rapid analysis of hundreds of molecular markers in the same set of specimens. Our detection of six gene amplifications as well as p53 and estrogen receptor expression in breast cancer demonstrates the power of this technique for defining new subgroups of tumors.

Keywords: Animals; Breast Neoplasms, genetics/metabolism/pathology; Cyclin D1, genetics/metabolism; Female; Genetic Techniques; Humans; Immunoenzyme Techniques; In Situ Hybridization, Fluorescence; Mice; Oncogene Proteins v-myb; Proto-Oncogene Proteins c-myc, genetics/metabolism; Rabbits; Receptor, erbB-2, genetics/metabolism; Receptors, Estrogen, genetics/metabolism; Retroviridae Proteins, Oncogenic, genetics/metabolism; Tumor Markers, Biological, genetics/metabolism; Tumor Suppressor Protein p53, genetics/metabolism

Activation of T cells requires recognition by T-cell receptors of specific peptides bound to major histocompatibility complex (MHC) molecules on the surface of either antigen-presenting or target cells. These peptides, T-cell epitopes, have potential therapeutic applications, such as for use as vaccines. Their identification, however, usually requires that multiple overlapping synthetic peptides encompassing a protein antigen be assayed, which in humans, is limited by volume of donor blood. T-cell epitopes are a subset of peptides that bind to MHC molecules. We use an artificial neural network (ANN) model trained to predict peptides that bind to the MHC class II molecule HLA-DR4(*0401). Binding prediction facilitates identification of T-cell epitopes in tyrosine phosphatase IA-2, an autoantigen in DR4-associated type1 diabetes. Synthetic peptides encompassing IA-2 were tested experimentally for DR4 binding and T-cell proliferation in humans at risk for diabetes. ANN-based binding prediction was sensitive and specific, and reduced the number of peptides required for T-cell assay by more than half, with only a minor loss of epitopes. This strategy could expedite identification of candidate T-cell epitopes in diverse diseases.

Keywords: immunoinformatics

This article shows a relationship between two different approximation techniques: the support vector machines (SVM), proposed by V. Vapnik (1995) and a sparse approximation scheme that resembles the basis pursuit denoising algorithm (Chen, 1995; Chen, Donoho, and Saunders, 1995). SVM is a technique that can be derived from the structural risk minimization principle (Vapnik, 1982) and can be used to estimate the parameters of several different approximation schemes, including radial basis functions, algebraic and trigonometric polynomials, B-splines, and some forms of multilayer perceptrons. Basis pursuit denoising is a sparse approximation technique in which a function is reconstructed by using a small number of basis functions chosen from a large set (the dictionary). We show that if the data are noiseless, the modified version of basis pursuit denoising proposed in this article is equivalent to SVM in the following sense: if applied to the same data set, the two techniques give the same solution, which is obtained by solving the same quadratic programming problem. In the appendix, we present a derivation of the SVM technique in one framework of regularization theory, rather than statistical learning theory, establishing a connection between SVM, sparse approximation, and regularization theory.

Keywords: Algorithms, Automated, Biometry, Computers, DNA, Databases, Factual, Fungal, Fungal Proteins, GTP-Binding Proteins, Gene Expression, Genes, Learning, Markov Chains, Models, Neural Networks (Computer), Neurological, Non-P.H.S., Non-U.S. Gov't, Nucleic Acid Hybridization, Open Reading Frames, P.H.S., Pattern Recognition, Protein, Protein Structure, Proteins, Reproducibility of Results, Research Support, Saccharomyces cerevisiae, Sequence Alignment, Sequence Analysis, Software, Statistical, Tertiary, U.S. Gov't, 9698353

Experimental introduction of RNA into cells can be used in certain biological systems to interfere with the function of an endogenous gene. Such effects have been proposed to result from a simple antisense mechanism that depends on hybridization between the injected RNA and endogenous messenger RNA transcripts. RNA interference has been used in the nematode Caenorhabditis elegans to manipulate gene expression. Here we investigate the requirements for structure and delivery of the interfering RNA. To our surprise, we found that double-stranded RNA was substantially more effective at producing interference than was either strand individually. After injection into adult animals, purified single strands had at most a modest effect, whereas double-stranded mixtures caused potent and specific interference. The effects of this interference were evident in both the injected animals and their progeny. Only a few molecules of injected double-stranded RNA were required per affected cell, arguing against stochiometric interference with endogenous mRNA and suggesting that there could be a catalytic or amplification component in the interference process.

Keywords: sirna

Keywords: chemoinformatics

The relationship between prediction and data compression can be extended to universal prediction schemes and universal data compression. Previous work shows that minimizing the sequential squared prediction error for individual sequences can be achieved using the same strategies which minimize the sequential code length for data compression of individual sequences. Defining a ?probability? as an exponential function of sequential loss, results from universal data compression can be used to develop universal linear prediction algorithms. Specifically, we present an algorithm for linear prediction of individual sequences which is twice-universal, over parameters and model orders

Keywords: information-theory

There have been many randomised trials of adjuvant prolonged polychemotherapy among women with early breast cancer, and an updated overview of their results is presented.In 1995, information was sought on each woman in any randomised trial that began before 1990 and involved treatment groups that differed only with respect to the chemotherapy regimens that were being compared. Analyses involved about 18,000 women in 47 trials of prolonged polychemotherapy versus no chemotherapy, about 6000 in 11 trials of longer versus shorter polychemotherapy, and about 6000 in 11 trials of anthracycline-containing regimens versus CMF (cyclophosphamide, methotrexate, and fluorouracil).For recurrence, polychemotherapy produced substantial and highly significant proportional reductions both among women aged under 50 at randomisation (35% [SD 4] reduction; 2p<0.00001) and among those aged 50-69 (20% [SD 3] reduction; 2p<0.00001); few women aged 70 or over had been studied. For mortality, the reductions were also significant both among women aged under 50 (27% [SD 5] reduction; 2p<0.00001) and among those aged 50-69 (11% [SD 3] reduction; 2p=0.0001). The recurrence reductions emerged chiefly during the first 5 years of follow-up, whereas the difference in survival grew throughout the first 10 years. After standardisation for age and time since randomisation, the proportional reductions in risk were similar for women with node-negative and node-positive disease. Applying the proportional mortality reduction observed in all women aged under 50 at randomisation would typically change a 10-year survival of 71% for those with node-negative disease to 78% (an absolute benefit of 7%), and of 42% for those with node-positive disease to 53% (an absolute benefit of 11%). The smaller proportional mortality reduction observed in all women aged 50-69 at randomisation would translate into smaller absolute benefits, changing a 10-year survival of 67% for those with node-negative disease to 69% (an absolute gain of 2%) and of 46% for those with node-positive disease to 49% (an absolute gain of 3%). The age-specific benefits of polychemotherapy appeared to be largely irrespective of menopausal status at presentation, oestrogen receptor status of the primary tumour, and of whether adjuvant tamoxifen had been given. In terms of other outcomes, there was a reduction of about one-fifth (2p=0.05) in contralateral breast cancer, which has already been included in the analyses of recurrence, and no apparent adverse effect on deaths from causes other than breast cancer (death rate ratio 0.89 [SD 0.09]). The directly randomised comparisons of longer versus shorter durations of polychemotherapy did not indicate any survival advantage with the use of more than about 3-6 months of polychemotherapy. By contrast, directly randomised comparisons did suggest that, compared with CMF alone, the anthracycline-containing regimens studied produced somewhat greater effects on recurrence (2p=0.006) and mortality (69% vs 72% 5-year survival; log-rank 2p=0.02). But this comparison is one of many that could have been selected for emphasis, the 99% CI reaches zero, and the results of several of the relevant trials are not yet available.Some months of adjuvant polychemotherapy (eg, with CMF or an anthracycline-containing regimen) typically produces an absolute improvement of about 7-11% in 10-year survival for women aged under 50 at presentation with early breast cancer, and of about 2-3% for those aged 50-69 (unless their prognosis is likely to be extremely good even without such treatment). Treatment decisions involve consideration not only of improvements in cancer recurrence and survival but also of adverse side-effects of treatment, and this report makes no recommendations as to who should or should not be treated.

Keywords: Adult; Aged; Antineoplastic Combined Chemotherapy Protocols, therapeutic use; Breast Neoplasms, chemistry/drug therapy/mortality; Chemotherapy, Adjuvant; Drug Administration Schedule; Female; Humans; Lymphatic Metastasis; Menopause; Middle Aged; Neoplasm Recurrence, Local; Randomized Controlled Trials as Topic; Receptors, Estrogen, analysis; Tamoxifen, administration /&/ dosage

Keywords: chemoinformatics

Keywords: chemoinformatics

We review the principles of minimum description length and stochastic complexity as used in data compression and statistical modeling. Stochastic complexity is formulated as the solution to optimum universal coding problems extending Shannon's basic source coding theorem. The normalized maximized likelihood, mixture, and predictive codings are each shown to achieve the stochastic complexity to within asymptotically vanishing terms. We assess the performance of the minimum description length criterion both from the vantage point of quality of data compression and accuracy of statistical inference. Context tree modeling, density estimation, and model selection in Gaussian linear regression serve as examples

Keywords: information-theory

Support vector machines (SVMs) perform pattern recognition between two point classes by finding a decision surface determined by certain points of the training set, termed support vectors (SV). This surface, which in some feature space of possibly infinite dimension can be regarded as a hyperplane, is obtained from the solution of a problem of quadratic programming that depends on a regularization parameter. In this article, we study some mathematical properties of support vectors and show that the decision surface can be written as the sum of two orthogonal terms, the first depending on only the margin vectors (which are SVs lying on the margin), the second proportional to the regularization parameter. For almost all values of the parameter, this enables us to predict how the decision surface varies for small parameter changes. In the special but important case of feature space of finite dimension m, we also show that m + 1 SVs are usually sufficient to determine the decision surface fully. For relatively small m, this latter result leads to a consistent reduction of the SV number.

Keywords: Algorithms, Artificial Intelligence, Automated, Biometry, Computers, DNA, Databases, Factual, Fungal, Fungal Proteins, GTP-Binding Proteins, Gene Expression, Genes, Learning, Linear Models, Markov Chains, Mathematics, Models, Neural Networks (Computer), Neurological, Non-P.H.S., Non-U.S. Gov't, Nonlinear Dynamics, Nucleic Acid Hybridization, Open Reading Frames, P.H.S., Pattern Recognition, Protein, Protein Structure, Proteins, Reproducibility of Results, Research Support, Saccharomyces cerevisiae, Sequence Alignment, Sequence Analysis, Software, Statistical, Tertiary, U.S. Gov't, 9573414

Whole-genome mRNA quantitation can be used to identify the genes that are most responsive to environmental or genotypic change. By searching for mutually similar DNA elements among the upstream non-coding DNA sequences of these genes, we can identify candidate regulatory motifs and corresponding candidate sets of coregulated genes. We have tested this strategy by applying it to three extensively studied regulatory systems in the yeast Saccharomyces cerevisiae: galactose response, heat shock, and mating type. Galactose-response data yielded the known binding site of Gal4, and six of nine genes known to be induced by galactose. Heat shock data yielded the cell-cycle activation motif, which is known to mediate cell-cycle dependent activation, and a set of genes coding for all four nucleosomal proteins. Mating type alpha and a data yielded all of the four relevant DNA motifs and most of the known a- and alpha-specific genes.

Keywords: bioinformatics, genome-wide, tfs

Of the many thousands of peptides encoded by a complex foreign antigen that can potentially be presented to CD8+ T cells (TCD8+), only a small fraction induce measurable responses in association with any given major histocompatibility complex class I allele. To design vaccines that elicit optimal TCD8+ responses, a thorough understanding of this phenomenon, known as immunodominance, is imperative. Here we review recent progress in unraveling the molecular and cellular basis for immunodominance. Of foremost importance is peptide binding to class I molecules; only approximately 1/200 of potential determinants bind at greater than the threshold affinity (Kd > 500 nM) associated with immunogenicity. Limitations in the TCD8+ repertoire render approximately half of these peptides nonimmunogenic, and inefficient antigen processing further thins the ranks by approximately four fifths. As a result, only approximately 1/2000 of the peptides in a foreign antigen expressed by an appropriate antigen presenting cell achieve immunodominant status with a given class I allele. A roughly equal fraction of peptides have subdominant status, i.e. they induce weak-to-nondetectable primary TCD8+ responses in the context of their natural antigen. Subdominant determinants may be expressed at or above levels of immunodominant determinants, at least on antigen presenting cells in vitro. The immunogenicity of subdominant determinants is often limited by immunodomination: suppression mediated by TCD8+ specific for immunodominant determinants. Immunodomination is a central feature of TCD8+ responses, as it even occurs among clones responding to the same immunodominant determinant. Little is known about how immunodominant and subdominant determinants are distinguished by the TCD8+ repertoire, or how (and why) immunodomination occurs, but new tools are available to address these questions.

Keywords: immunoinformatics

The adoptive transfer of tumor-infiltrating lymphocytes along with interleukin 2 into autologous patients resulted in the objective regression of tumor in about 30% of patients with melanoma, indicating that these T cells play a role in tumor rejection. To understand the molecular basis of the T cell-cancer cell interaction we and others started to search for tumor antigens expressed on cancer cells recognized by T cells. This led to the identification of several major histocompatibility complex (MHC) class I restricted tumor antigens. These tumor antigens have been classified into several categories: tissue-specific differentiation antigens, tumor-specific shared antigens, and tumor-specific unique antigens. Because CD4+ T cells play a central role in orchestrating the host immune response against cancer, infectious diseases, and autoimmune diseases, a novel genetic approach has recently been developed to identify these MHC class II restricted tumor antigens. The identification of both MHC class I and II restricted tumor antigens provides new opportunities for the development of therapeutic strategies against cancer. This review summarizes the current status of tumor antigens and their potential applications to cancer treatment.

Keywords: immunoinformatics

Double-stranded RNA (dsRNA) directs gene-specific, post-transcriptional silencing in many organisms, including vertebrates, and has provided a new tool for studying gene function. The biochemical mechanisms underlying this dsRNA interference (RNAi) are unknown. Here we report the development of a cell-free system from syncytial blastoderm Drosophila embryos that recapitulates many of the features of RNAi. The interference observed in this reaction is sequence specific, is promoted by dsRNA but not single-stranded RNA, functions by specific mRNA degradation, and requires a minimum length of dsRNA. Furthermore, preincubation of dsRNA potentiates its activity. These results demonstrate that RNAi can be mediated by sequence-specific processes in soluble reactions.

Keywords: csbcbook

Array technologies have made it straightforward to monitor simultaneously the expression pattern of thousands of genes. The challenge now is to interpret such massive data sets. The first step is to extract the fundamental patterns of gene expression inherent in the data. This paper describes the application of self-organizing maps, a type of mathematical cluster analysis that is particularly well suited for recognizing and classifying features in complex, multidimensional data. The method has been implemented in a publicly available computer package, GENECLUSTER, that performs the analytical calculations and provides easy data visualization. To illustrate the value of such analysis, the approach is applied to hematopoietic differentiation in four well studied models (HL-60, U937, Jurkat, and NB4 cells). Expression patterns of some 6,000 human genes were assayed, and an online database was created. GENECLUSTER was used to organize the genes into biologically relevant clusters that suggest novel hypotheses about hematopoietic differentiation-for example, highlighting certain genes and pathways involved in "differentiation therapy" used in the treatment of acute promyelocytic leukemia.

Studies into posttranslational modifications of histones, notably acetylation, have yielded important insights into the dynamic nature of chromatin structure and its fundamental role in gene expression. The roles of other covalent histone modifications remain poorly understood. To gain further insight into histone methylation, we investigated its occurrence and pattern of site utilization in Tetrahymena, yeast, and human HeLa cells. In Tetrahymena, transcriptionally active macronuclei, but not transcriptionally inert micronuclei, contain a robust histone methyltransferase activity that is highly selective for H3. Microsequence analyses of H3 from Tetrahymena, yeast, and HeLa cells indicate that lysine 4 is a highly conserved site of methylation, which to date, is the major site detected in Tetrahymena and yeast. These data document a nonrandom pattern of H3 methylation that does not overlap with known acetylation sites in this histone. In as much as H3 methylation at lysine 4 appears to be specific to macronuclei in Tetrahymena, we suggest that this modification pattern plays a facilitatory role in the transcription process in a manner that remains to be determined. Consistent with this possibility, H3 methylation in yeast occurs preferentially in a subpopulation of H3 that is preferentially acetylated.

Keywords: Acetyltransferases, metabolism; Amino Acid Sequence; Animals; Cell Nucleus, metabolism; Hela Cells; Histone Acetyltransferases; Histone-Lysine N-Methyltransferase; Histones, metabolism; Humans; Lysine, analogs /&/ derivatives/metabolism; Methylation; Methyltransferases, metabolism; Molecular Sequence Data; Protein Methyltransferases; Protein Processing, Post-Translational; Saccharomyces cerevisiae Proteins; Species Specificity; Tetrahymena thermophila; Transcription, Genetic; Yeasts

Keywords: csbcbook, csbcbook-ch2

Herein, we review the epitope approach to vaccine development, and discuss how knowledge of HLA supertypes might be used as a tool in the development of such vaccines. After reviewing the main structural features of the A2-, A3-, B7-, and B44- supertype alleles, and biological data demonstrating their immunological relevance, we analyze the frequency at which these supertype alleles are expressed in various ethnicities and discuss the relevance of those observations to vaccine development. Next, the existence of five new supertypes (A1, A24, B27, B58, and B62) is reported. As a result, it is possible to account for the predominance of all known HLA class I with only nine main functional binding specificities. The practical implications of this finding, as well as its relevance to understanding the functional implication of MHC polymorphism in humans, are discussed.

Keywords: Alleles; Amino Acid Sequence; Animals; Epitopes; HLA-A Antigens; HLA-B Antigens; Histocompatibility Antigens Class I; Humans; Molecular Sequence Data; Polymorphism, Genetic

We have developed a generic procedure to purify proteins expressed at their natural level under native conditions using a novel tandem affinity purification (TAP) tag. The TAP tag allows the rapid purification of complexes from a relatively small number of cells without prior knowledge of the complex composition, activity, or function. Combined with mass spectrometry, the TAP strategy allows for the identification of proteins interacting with a given target protein. The TAP method has been tested in yeast but should be applicable to other cells or organisms.

Keywords: Affinity Labels; Amino Acid Sequence; Electrophoresis, Polyacrylamide Gel; Methods; Molecular Sequence Data; Proteins; Proteome

The first version of the major histocompatibility complex (MHC) databank SYFPEITHI: database for MHC ligands and peptide motifs, is now available to the general public. It contains a collection of MHC class I and class II ligands and peptide motifs of humans and other species, such as apes, cattle, chicken, and mouse, for example, and is continuously updated. All motifs currently available are accessible as individual entries. Searches for MHC alleles, MHC motifs, natural ligands, T-cell epitopes, source proteins/organisms and references are possible. Hyperlinks to the EMBL and PubMed databases are included. In addition, ligand predictions are available for a number of MHC allelic products. The database content is restricted to published data only.

Keywords: Amino Acid Motifs; Amino Acid Sequence; Animals; Databases, Factual; Humans; Internet; Ligan; Major Histocompatibility Complex; Molecular Sequence Data; Research Support, Non-U.S. Gov't; Sequence Homology, Amino Acid; ds

Keywords: csbcbook, csbcbook-ch2

Keywords: kernel-theory

cDNA microarrays and a clustering algorithm were used to identify patterns of gene expression in human mammary epithelial cells growing in culture and in primary human breast tumors. Clusters of coexpressed genes identified through manipulations of mammary epithelial cells in vitro also showed consistent patterns of variation in expression among breast tumor samples. By using immunohistochemistry with antibodies against proteins encoded by a particular gene in a cluster, the identity of the cell type within the tumor specimen that contributed the observed gene expression pattern could be determined. Clusters of genes with coherent expression patterns in cultured cells and in the breast tumors samples could be related to specific features of biological variation among the samples. Two such clusters were found to have patterns that correlated with variation in cell proliferation rates and with activation of the IFN-regulated signal transduction pathway, respectively. Clusters of genes expressed by stromal cells and lymphocytes in the breast tumors also were identified in this analysis. These results support the feasibility and usefulness of this systematic approach to studying variation in gene expression patterns in human cancers as a means to dissect and classify solid tumors.

Keywords: csbcbook, csbcbook-ch3

Keywords: biogm

A new method of rapid pharmacophore fingerprinting (PharmPrint method) has been developed. A basis set of 10,549 three-point pharmacophores has been constructed by enumerating several distance ranges and pharmacophoric features. Software has been developed to assign pharmacophoric types to atoms in chemical structures, generate multiple conformations, and construct the binary fingerprint according to the pharmacophores that result. The fingerprint is used as a descriptor for developing a quantitative structure-activity relationship (QSAR) model using partial least squares. An example is given using sets of ligands for the estrogen receptor (ER). The result is compared with previously published results on the same data to show the superiority of a full 3D, conformationally flexible approach. The QSAR model can be readily interpreted in structural/chemical terms. Further examples are given using binary activity data and some of our novel in-house compounds, which show the value of the model when crossing compound classes.

Keywords: Chemistry, Combinatorial Chemistry Techniques, Drug Design, Drug Evaluation, Estradiol Congeners, Estrogen, Least-Squares Analysis, Ligands, Models, Molecular, Pharmaceutical, Preclinical, Receptors, Software, Structure-Activity Relationship, 10361729

The performance of two important 2D and 3D molecular descriptors for rational design to maximize the structural diversity of databases is investigated in this publication. Those methods are based either on a 2D description using a binary fingerprint, which accounts for the absence or presence of molecular fragments, or a 3D description based on the geometry of pharmacophoric features encoded in a fingerprint (pharmacophoric definition triplets, PDTs). Both descriptors in combination with maximum dissimilarity selections, complete linkage hierarchical cluster analysis, or sequential dissimilarity selections were compared to random subsets as reference. This comparison is based on their ability to cover representative biological classes from parent databases (coverage analysis) and the degree of separation between active and inactive compounds for a biological target from hierarchical clustering (cluster separation analysis). While the similarity coefficients (Tanimoto, cosine) show only a minor influence, the number of conformations to generate the 3D PDT fingerprint lead to remarkably different results. PDT fingerprints derived from a lower number of conformers perform significantly better, but they are not comparable to a 2D fingerprint-based design. When 2D and 3D descriptors are combined with weighting factors > 0.5 for 2D fingerprints, a significant improvement of coverage and cluster separation results is observed for a small number of PDT conformers and medium sized subsets. Some combined descriptors outperform 2D fingerprints, but not for all subset populations. Applying sequential dissimilarity selection to PDT descriptors reveals that its performance is dependent on the initial ordering of compounds, while presorting according to 2D fingerprint diversity does not improve results. Finally the relationship between biological activity and similarity was investigated, showing that PDTs quantify smaller structural differences due to the large number of bits in the fingerprint.

Keywords: chemoinformatics

A new 4-point pharmacophore method for molecular similarity and diversity that rapidly calculates all potential pharmacophores/pharmacophoric shapes for a molecule or a protein site is described. The method, an extension to the ChemDiverse/Chem-X software (Oxford Molecular, Oxford, England), has also been customized to enable a new internally referenced measure of pharmacophore diversity. The "privileged" substructure concept for the design of high-affinity ligands is presented, and an example of this new method is described for the design of combinatorial libraries for 7-transmembrane G-protein-coupled receptor targets, where "privileged" substructures are used as special features to internally reference the pharmacophoric shapes. Up to 7 features and 15 distance ranges are considered, giving up to 350 million potential 4-point 3D pharmacophores/molecule. The resultant pharmacophore "key" ("fingerprint") serves as a powerful measure for diversity or similarity, calculable for both a ligand and a protein site, and provides a consistent frame of reference for comparing molecules, sets of molecules, and protein sites. Explicit "on-the-fly" conformational sampling is performed for a molecule to enable the calculation of all geometries accessible for all combinations of four features (i.e., 4-point pharmacophores) at any desired sampling resolution. For a protein site, complementary site points to groups displayed in the site are generated and all combinations of four site points are considered. In this paper we report (i) the details of our customized implementation of the method and its modification to systematically measure 4-point pharmacophores relative to a "special" substructure of interest present in the molecules under study; (ii) comparisons of 3- and 4-point pharmacophore methods, highlighting the much increased resolution of the 4-point method; (iii) applications of the 4-point potential pharmacophore descriptors as a new measure of molecular similarity and diversity and for the design of focused/biased combinatorial libraries.

Is perception of the whole based on perception of its parts? There is psychological and physiological evidence for parts-based representations in the brain, and certain computational theories of object recognition rely on such representations. But little is known about how brains or computers might learn the parts of objects. Here we demonstrate an algorithm for non-negative matrix factorization that is able to learn parts of faces and semantic features of text. This is in contrast to other methods, such as principal components analysis and vector quantization, that learn holistic, not parts-based, representations. Non-negative matrix factorization is distinguished from the other methods by its use of non-negativity constraints. These constraints lead to a parts-based representation because they allow only additive, not subtractive, combinations. When non-negative matrix factorization is implemented as a neural network, parts-based representations emerge by virtue of two properties: the firing rates of neurons are never negative and synaptic strengths do not change sign.

Keywords: Animals; DNA, analysis; Gene Expression; Genetic Variation; Genome; Humans; Molecular Probe Techniques, trends; Oligonucleotide Array Sequence Analysis, methods; RNA, Messenger, analysis; Saccharomyces cerevisiae, genetics

Keywords: PUlearning

Keywords: biosvm

Keywords: biosvm

Keywords: glycans

Keywords: biogm

We introduce a new method of constructing kernels on sets whose elements are discrete structures like strings, trees and graphs. The method can be applied iteratively to build a kernel on a infinite set from kernels involving generators of the set. The family of kernels generated generalizes the family of radial basis kernels. It can also be used to define kernels in the form of joint Gibbs probability distributions. Kernels can be built from hidden Markov random fields, generalized regular expressions, pair-HMMs, or ANOVA decompositions. Uses of the method lead to open problems involving the theory of infinitely divisible positive definite functions. Fundamentals of this theory and the theory of reproducing kernel Hilbert spaces are reviewed and applied in establishing the validity of the method.

Keywords: biosvm

Cellular functions, such as signal transmission, are carried out by 'modules' made up of many species of interacting molecules. Understanding how modules work has depended on combining phenomenological analysis with molecular studies. General principles that govern the structure and behaviour of modules may be discovered with help from synthetic sciences such as engineering and computer science, from stronger interactions between experiment and theory in cell biology, and from an appreciation of evolutionary constraints.

Keywords: Action Potentials; Biological Evolution; Forecasting; Models, Biological; Molecular Biology, trends

Library chemistry and high-throughput screening require greater use of chemoinformatics to increase their effectiveness. Recent advances in chemoinformatics include new molecular descriptors and pharmacophore techniques, statistical tools and their applications. Visualisation methods and hardware development are also opening new opportunities. The advent of a chemically aware web language and cross-platform working is ensuring that chemoinformatics methods are becoming available to all chemists in a more appropriate manner. Much time will continue to be wasted with incompatible file types without internationally agreed standards.

Keywords: Chemistry, Computers, Drug Design, Information Science, Software, 10419846

We present a simple randomized procedure for the prediction of a binary sequence. The algorithm uses ideas from previous developments of the theory of the prediction of individual sequences. We show that if the sequence is a realization of a stationary and ergodic random process then the average number of mistakes converges, almost surely, to that of the optimum, given by the Bayes predictor. The desirable finite-sample properties of the predictor are illustrated by its performance for Markov processes. In such cases the predictor exhibits near-optimal behavior even without knowing the order of the Markov process. Prediction with side information is also considered

Keywords: information-theory

We describe an approach for the accurate quantification and concurrent sequence identification of the individual proteins within complex mixtures. The method is based on a class of new chemical reagents termed isotope-coded affinity tags (ICATs) and tandem mass spectrometry. Using this strategy, we compared protein expression in the yeast Saccharomyces cerevisiae, using either ethanol or galactose as a carbon source. The measured differences in protein expression correlated with known yeast metabolic function under glucose-repressed conditions. The method is redundant if multiple cysteinyl residues are present, and the relative quantification is highly accurate because it is based on stable isotope dilution techniques. The ICAT approach should provide a widely applicable means to compare quantitatively global protein expression in cells and tissues.

Keywords: Affinity Labels; Amino Acid Sequence; Chromatography, Liquid; Isotope Labeling; Mass Spectrometry; Proteins

Although cancer classification has improved over the past 30 years, there has been no general approach for identifying new cancer classes (class discovery) or for assigning tumors to known classes (class prediction). Here, a generic approach to cancer classification based on gene expression monitoring by DNA microarrays is described and applied to human acute leukemias as a test case. A class discovery procedure automatically discovered the distinction between acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL) without previous knowledge of these classes. An automatically derived class predictor was able to determine the class of new leukemia cases. The results demonstrate the feasibility of cancer classification based solely on gene expression moni- toring and suggest a general strategy for discovering and predicting cancer classes for other types of cancer, independent of previous biological knowledge.

Keywords: csbcbook, csbcbook-ch3, csbcbook-ch4

DNA microarrays can be used to measure the expression patterns of thousands of genes in parallel, generating clues to gene function that can help to identify appropriate targets for therapeutic intervention. They can also be used to monitor changes in gene expression in response to drug treatments. Here, we discuss the different ways in which microarray analysis is likely to affect drug discovery.

Keywords: Agricultural, Alleles, Alternaria, Amino Acid, Amino Acid Chloromethyl Ketones, Amino Acid Sequence, Animal, Animals, Apoptosis, Asthma, Bacteria, Base Sequence, Binding Sites, Biotechnology, Blotting, Bone Density, Bone Matrix, Bone and Bones, CCR5, Camptothecin, Caspases, Cathepsins, Cell Surface, Central America, Chloroplast, Chondrocytes, Chromosome Mapping, Chromosomes, Cloning, Cluster Analysis, Collagen, Comparative Study, Coumarins, Crops, Crystallography, DNA, DNA Primers, Dipeptides, Disease, Disease Models, Drug Design, Drug Evaluation, Drug Industry, Enzyme Activation, Enzyme Inhibitors, Escherichia coli, Evolution, Exons, Expressed Sequence Tags, Female, Fetus, Fluorescent Dyes, Food Microbiology, Founder Effect, GTP-Binding Proteins, Gene Expression, Gene Frequency, Gene Library, Genes, Genetic, Genetic Predisposition to Disease, Genome, Geography, Growth Plate, Haplotypes, Hordeum, Human, Humans, Inclusion Bodies, Injections, Intraperitoneal, Introns, Isatin, Knockout, Male, Membrane Proteins, Messenger, Mice, Models, Molecular, Molecular Sequence Data, Molecular Structure, Mutation, Mycotoxins, Neutrophils, Non-U.S. Gov't, Northern, Oligonucleotide Array Sequence Analysis, Osteoarthritis, Osteochondrodysplasias, Osteoclasts, Osteopetrosis, Pair 15, Phaseolus, Polymorphism, Preclinical, Pregnancy, Promoter Regions (Genetics), Protein Precursors, Proteomics, RNA, Receptors, Recombinant Fusion Proteins, Recombinant Proteins, Research Support, Restriction Fragment Length, Ribosomal Proteins, Sequence Alignment, Sequence Analysis, Sequence Homology, South America, Species Specificity, Splenomegaly, Sulfonamides, Synteny, Tissue Distribution, Transcription, Trichothecenes, X-Ray, 9915501

We propose a differential equation model for gene expression and provide two methods to construct the model from a set of temporal data. We model both transcription and translation by kinetic equations with feedback loops from translation products to transcription. Degradation of proteins and mRNAs is also incorporated. We study two methods to construct the model from experimental data: Minimum Weight Solutions to Linear Equations (MWSLE), which determines the regulation by solving under-determined linear equations, and Fourier Transform for Stable Systems (FTSS), which refines the model with cell cycle constraints. The results suggest that a minor set of temporal data may be sufficient to construct the model at the genome level. We also give a comprehensive discussion of other extended models: the RNA Model, the Protein Model, and the Time Delay Model.

Among membrane-bound receptors, the G protein-coupled receptors (GPCRs) are certainly the most diverse. They have been very successful during evolution, being capable of transducing messages as different as photons, organic odorants, nucleotides, nucleosides, peptides, lipids and proteins. Indirect studies, as well as two-dimensional crystallization of rhodopsin, have led to a useful model of a common 'central core', composed of seven transmembrane helical domains, and its structural modifications during activation. There are at least six families of GPCRs showing no sequence similarity. They use an amazing number of different domains both to bind their ligands and to activate G proteins. The fine-tuning of their coupling to G proteins is regulated by splicing, RNA editing and phosphorylation. Some GPCRs have been found to form either homo- or heterodimers with a structurally different GPCR, but also with membrane-bound proteins having one transmembrane domain such as nina-A, odr-4 or RAMP, the latter being involved in their targeting, function and pharmacology. Finally, some GPCRs are unfaithful to G proteins and interact directly, via their C-terminal domain, with proteins containing PDZ and Enabled/VASP homology (EVH)-like domains.

Keywords: chemogenomics

Keywords: csbcbook

Systems as diverse as genetic networks or the World Wide Web are best described as networks with complex topology. A common property of many large networks is that the vertex connectivities follow a scale-free power-law distribution. This feature was found to be a consequence of two generic mechanisms: (i) networks expand continuously by the addition of new vertices, and (ii) new vertices attach preferentially to sites that are already well connected. A model based on these two ingredients reproduces the observed stationary scale-free distributions, which indicates that the development of large networks is governed by robust self-organizing phenomena that go beyond the particulars of the individual systems.

We propose a method of modifying a kernel function to improve the performance of a support vector machine classifier. This is based on the structure of the Riemannian geometry induced by the kernel function. The idea is to enlarge the spatial resolution around the separating boundary surface, by a conformal mapping, such that the separability between classes is increased. Examples are given specifically for modifying Gaussian Radial Basis Function kernels. Simulation results for both artificial and real data show remarkable improvement of generalization errors, supporting our idea.

A detailed model mechanism for the G1/S transition in the mammalian cell cycle is presented and analysed by computer simulation to investigate whether the kinetic origins of the restriction point (R-point) can be identified. The R-point occurs in mid-to-late G1 phase and marks the transition between mitogen-dependent to mitogen-independent progression of the cell cycle. For purposes of computer simulations, the R-point is defined as the first point in time after mitosis where cutting off mitogen stimulation does not prevent the cell reaching the threshold activity of cyclin-E/cdk2 required for entry into S phase. The key components of the network that generate a dynamic switching behaviour associated with the R-point include a positive feedback loop between cyclin-E/cdk2 and Cdc25A, along with the mutually negative interaction between the cdk inhibitor p27Kip1 and cyclin-E/cdk2. Simulations of the passage through the R-point were carried out and the factors affecting the position of the R-point in G1 are determined. The detailed model also shows various points in the network where the activation of cyclin-E/cdk2 can be initiated with or without the involvement of the retinoblastoma protein.

Keywords: csbcbook

A detailed model of the G(2) DNA damage checkpoint (G2DDC) system is presented that includes complex regulatory networks of the mitotic kinase Cdc2, phosphatase Cdc25, Wee1 kinase, and damage signal transduction pathways involving Chk1 and p53. Assumptions on the kinetic equations of the G2DDC are made, and computer simulations are carried out to demonstrate how the various subsystems operate to delay or arrest cell cycle progression. The detailed model could be used to explain various experiments relevant to G2DDC reported recently, including the nuclear export of 14-3-3-bound Cdc25, the down-regulation of cyclin B1 expression by p53, the effect of Chk1 and p53 on Cdc25 levels, and Wee1 degradation. It also is shown that, under certain conditions, p53 is necessary to sustain a G(2) arrest.

Keywords: csbcbook

The G2-M checkpoint in the cell cycle is identified with a set of phosphorylation-dephosphorylation (PD) cycles involving Cdc25 and the maturation-promoting factor (MPF); these PD cycles are coupled in a way that generates an instability. This instability arises out of a transcritical bifurcation which could be exploited by the G2 DNA damage checkpoint pathway in order to arrest or delay entry into mitosis. The coupling between PD cycles involving Wee1 and MPF does not lead to an instability and therefore Wee1 may not be a crucial target of the checkpoint pathway. A set of PD cycles exhibiting transcritical bifurcation also possesses the integrative ability of a checkpoint for 'checking' that prerequisites are satisfied prior to the next cell cycle event. Such a set of coupled PD cycles is suggested to be a core mechanism of cell cycle checkpoints.

Keywords: csbcbook

DNA microarray technologies together with rapidly increasing genomic sequence information is leading to an explosion in available gene expression data. Currently there is a great need for efficient methods to analyze and visualize these massive data sets. A self-organizing map (SOM) is an unsupervised neural network learning algorithm which has been successfully used for the analysis and organization of large data files. We have here applied the SOM algorithm to analyze published data of yeast gene expression and show that SOM is an excellent tool for the analysis and visualization of gene expression profiles.

An improved dynamic programming algorithm is reported for RNA secondary structure prediction by free energy minimization. Thermodynamic parameters for the stabilities of secondary structure motifs are revised to include expanded sequence dependence as revealed by recent experiments. Additional algorithmic improvements include reduced search time and storage for multibranch loop free energies and improved imposition of folding constraints. An extended database of 151,503 nt in 955 structures? determined by comparative sequence analysis was assembled to allow optimization of parameters not based on experiments and to test the accuracy of the algorithm. On average, the predicted lowest free energy structure contains 73 % of known base-pairs when domains of fewer than 700 nt are folded; this compares with 64 % accuracy for previous versions of the algorithm and parameters. For a given sequence, a set of 750 generated structures contains one structure that, on average, has 86 % of known base-pairs. Experimental constraints, derived from enzymatic and flavin mononucleotide cleavage, improve the accuracy of structure predictions.

Keywords: 16S, 23S, 5S, Affinity, Algorithms, Aluminum Silicates, Amino Acid, Amino Acid Sequence, Amyloidosis, Archaeal, Bacillus, Bacterial, Bacterial Proteins, Bacteriophage T4, Base Sequence, Chloroplast, Chromatography, Circular Dichroism, Comparative Study, Computational Biology, Databases, Electrophoresis, Entropy, Enzyme Stability, Escherichia coli, Factual, Fibroblast Growth Factor 2, Flavin Mononucleotide, Fluorescence, Genetic, Guanidine, Humans, Huntington Disease, Kinetics, Light, Models, Molecular Sequence Data, Non-P.H.S., Non-U.S. Gov't, Nucleic Acid Conformation, P.H.S., Peptides, Phylogeny, Polyacrylamide Gel, Predictive Value of Tests, Protein Binding, Protein Denaturation, Protein Folding, Protein Structure, RNA, Radiation, Recombinant Proteins, Research Support, Ribosomal, Scattering, Secondary, Sequence Homology, Solutions, Spectrometry, Statistical, Temperature, Thermodynamics, Time Factors, Trinucleotide Repeat Expansion, U.S. Gov't, alpha-Amylase, 10329189

An improved dynamic programming algorithm is reported for RNA secondary structure prediction by free energy minimization. Thermodynamic parameters for the stabilities of secondary structure motifs are revised to include expanded sequence dependence as revealed by recent experiments. Additional algorithmic improvements include reduced search time and storage for multibranch loop free energies and improved imposition of folding constraints. An extended database of 151,503 nt in 955 structures? determined by comparative sequence analysis was assembled to allow optimization of parameters not based on experiments and to test the accuracy of the algorithm. On average, the predicted lowest free energy structure contains 73 % of known base-pairs when domains of fewer than 700 nt are folded; this compares with 64 % accuracy for previous versions of the algorithm and parameters. For a given sequence, a set of 750 generated structures contains one structure that, on average, has 86 % of known base-pairs. Experimental constraints, derived from enzymatic and flavin mononucleotide cleavage, improve the accuracy of structure predictions.

Keywords: sirna

Motivation: In order to extract protein sequences from nucleotide sequences, it is an important step to recognize points at which regions start that code for proteins. These points are called translation initiation sites (TIS). Results: The task of finding TIS can be modeled as a classification problem. We demonstrate the applicability of support vector machines for this task, and show how to incorporate prior biological knowledge by engineering an appropriate kernel function. With the described techniques the recognition performance can be improved by 26 We provide evidence that existing related methods (e.g. ESTScan) could profit from advanced TIS recognition.

Keywords: biosvm

Double-stranded RNA (dsRNA) directs the sequence-specific degradation of mRNA through a process known as RNA interference (RNAi). Using a recently developed Drosophila in vitro system, we examined the molecular mechanism underlying RNAi. We find that RNAi is ATP dependent yet uncoupled from mRNA translation. During the RNAi reaction, both strands of the dsRNA are processed to RNA segments 21-23 nucleotides in length. Processing of the dsRNA to the small RNA fragments does not require the targeted mRNA. The mRNA is cleaved only within the region of identity with the dsRNA. Cleavage occurs at sites 21-23 nucleotides apart, the same interval observed for the dsRNA itself, suggesting that the 21-23 nucleotide fragments from the dsRNA are guiding mRNA cleavage.

Keywords: sirna

Many contemporary applications in computer-aided drug discovery and chemoinformatics depend on representations of molecules by descriptors that capture their structural characteristics and properties. Such applications include, among others, diversity analysis, library design, and virtual screening. Hundreds of molecular descriptors have been reported in the literature, ranging from simple bulk properties to elaborate three-dimensional formulations and complex molecular fingerprints, which sometimes consist of thousands of bit positions. Knowledge-based selection of descriptors that are suitable for specific applications is an important task in chemoinformatics research. If descriptors are to be selected on rational grounds, rather than guesses or chemical intuition, detailed evaluation of their performance is required. A number of studies have been reported that investigate the performance of molecular descriptors in specific applications and/or introduce novel types of descriptors. Progress made in this area is reviewed here in the context of other computational developments in combinatorial chemistry and compound screening.

Keywords: chemoinformatics

For problems of data compression, gambling, and prediction of individual sequences x1, ···, xn the following questions arise. Given a target family of probability mass functions p(x1, ···, x n|&thetas;), how do we choose a probability mass function q(x 1, ···, xn) so that it approximately minimizes the maximum regret/belowdisplayskip10ptminus6pt max (log1/q(x1, ···, xn)-log1/p(x1, ···, xn |&thetas;?)) and so that it achieves the best constant C in the asymptotics of the minimax regret, which is of the form (d/2)log(n/2?)+C+o(1), where d is the parameter dimension? Are there easily implementable strategies q that achieve those asymptotics? And how does the solution to the worst case sequence problem relate to the solution to the corresponding expectation version minq max 0 E0(log1/q(x1, ···, xn)-log1/p(x1, ···, xn|&thetas;))? In the discrete memoryless case, with a given alphabet of size m, the Bayes procedure with the Dirichlet(1/2, ···, 1/2) prior is asymptotically maximin. Simple modifications of it are shown to be asymptotically minimax. The best constant is Cm=log(?(1/2)m/(?(m/2)) which agrees with the logarithm of the integral of the square root of the determinant of the Fisher information. Moreover, our asymptotically optimal strategies for the worst case problem are also asymptotically optimal for the expectation version. Analogous conclusions are given for the case of prediction, gambling, and compression when, for each observation, one has access to side information from an alphabet of size k. In this setting the minimax regret is shown to be k(m-1)/2logn/2?k+kCm+o(1)

Keywords: information-theory

We are concerned with the rating of new documents that appear in a large database (MEDLINE) and are candidates for inclusion in a small specialty database (REBASE). The requirement is to rank the new documents as nearly in order of decreasing potential to be added to the smaller database as possible, so as to improve the coverage of the smaller database without increasing the effort of those who manage this specialty database. To perform this ranking task we have considered several machine learning approaches based on the naï ve Bayesian algorithm. We find that adaptive boosting outperforms naï ve Bayes, but that a new form of boosting which we term staged Bayesian retrieval outperforms adaptive boosting. Staged Bayesian retrieval involves two stages of Bayesian retrieval and we further find that if the second stage is replaced by a support vector machine we again obtain a significant improvement over the strictly Bayesian approach.

Keywords: Acute, Acute Disease, Adenocarcinoma, Algorithms, Amino Acid Sequence, Animals, Artificial Intelligence, Automated, B-Lymphocytes, Bacterial Proteins, Base Pair Mismatch, Base Sequence, Bayes Theorem, Binding Sites, Biological, Bone Marrow Cells, Brachyura, Cell Compartmentation, Chemistry, Child, Chromosome Aberrations, Classification, Codon, Colonic Neoplasms, Comparative Study, Computational Biology, Computer Simulation, Computer-Assisted, DNA, Data Interpretation, Databases, Decision Trees, Diabetes Mellitus, Diagnosis, Discriminant Analysis, Discrimination Learning, Electric Conductivity, Electrophysiology, Escherichia coli Proteins, Factual, Feedback, Female, Fungal, Gastric Emptying, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Markers, Genetic Predisposition to Disease, Genomics, Hemolysins, Humans, Indians, Information Storage and Retrieval, Initiator, Ion Channels, Kinetics, Leukemia, Likelihood Functions, Lipid Bilayers, Logistic Models, Lymphocytic, MEDLINE, Male, Markov Chains, Melanoma, Models, Molecular, Myeloid, Neoplasm, Neoplasms, Neoplastic, Neural Networks (Computer), Neurological, Nevus, Non-P.H.S., Non-U.S. Gov't, Nonlinear Dynamics, Normal Distribution, North American, Nucleic Acid Conformation, Oligonucleotide Array Sequence Analysis, Organ Specificity, Organelles, Ovarian Neoplasms, Ovary, P.H.S., Pattern Recognition, Physical, Pigmented, Predictive Value of Tests, Promoter Regions (Genetics), Protein Biosynthesis, Protein Folding, Protein Structure, Proteins, Proteome, RNA, Reproducibility of Results, Research Support, Saccharomyces cerevisiae, Secondary, Sensitivity and Specificity, Sequence Alignment, Sequence Analysis, Sex Characteristics, Skin Diseases, Skin Neoplasms, Skin Pigmentation, Software, Sound Spectrography, Statistical, Stomach Diseases, T-Lymphocytes, Thermodynamics, Transcription, Transcription Factors, Tumor Markers, Type 2, U.S. Gov't, Vertebrates, 11080018

Keywords: biosvm

We introduce the concept of span of support vectors (SV) and show that the generalization ability of support vector machines (SVM) depends on this new geometrical concept. We prove that the value of the span is always smaller (and can be much smaller) than the diameter of the smallest sphere containing the support vectors, used in previous bounds (Vapnik, 1998). We also demonstrate experimentally that the prediction of the test error given by the span is very accurate and has direct application in model selection (choice of the optimal parameters of the SVM).

Keywords: Automated, Learning, Models, Neural Networks (Computer), Neurological, Pattern Recognition, 10976137

Scientists working with large volumes of high-dimensional data, such as global climate patterns, stellar spectra, or human gene distributions, regularly confront the problem of dimensionality reduction: finding meaningful low-dimensional structures hidden in their high-dimensional observations. The human brain confronts the same problem in everyday perception, extracting from its high-dimensional sensory inputs-30,000 auditory nerve fibers or 10(6) optic nerve fibers-a manageably small number of perceptually relevant features. Here we describe an approach to solving dimensionality reduction problems that uses easily measured local metric information to learn the underlying global geometry of a data set. Unlike classical techniques such as principal component analysis (PCA) and multidimensional scaling (MDS), our approach is capable of discovering the nonlinear degrees of freedom that underlie complex natural observations, such as human handwriting or images of a face under different viewing conditions. In contrast to previous algorithms for nonlinear dimensionality reduction, ours efficiently computes a globally optimal solution, and, for an important class of data manifolds, is guaranteed to converge asymptotically to the true structure.

Keywords: dimred

Histone proteins and the nucleosomes they form with DNA are the fundamental building blocks of eukaryotic chromatin. A diverse array of post-translational modifications that often occur on tail domains of these proteins has been well documented. Although the function of these highly conserved modifications has remained elusive, converging biochemical and genetic evidence suggests functions in several chromatin-based processes. We propose that distinct histone modifications, on one or more tails, act sequentially or in combination to form a 'histone code' that is, read by other proteins to bring about distinct downstream events.

Keywords: Acetylation; Amino Acid Sequence; Animals; Chromatin, physiology; Histones, chemistry/metabolism/physiology; Humans; Lysine, physiology; Microtubules, physiology; Models, Biological; Molecular Sequence Data; Phosphorylation; Protein Processing, Post-Translational; Serine, metabolism

A set of linear pathways often does not capture the full range of behaviors of a metabolic network. The concept of 'elementary flux modes' provides a mathematical tool to define and comprehensively describe all metabolic routes that are both stoichiometrically and thermodynamically feasible for a group of enzymes. We have used this concept to analyze the interplay between the pentose phosphate pathway (PPP) and glycolysis. The set of elementary modes for this system involves conventional glycolysis, a futile cycle, all the modes of PPP function described in biochemistry textbooks, and additional modes that are a priori equally entitled to pathway status. Applications include maximizing product yield in amino acid and antibiotic synthesis, reconstruction and consistency checks of metabolism from genome data, analysis of enzyme deficiencies, and drug target identification in metabolic networks.

Specific binding of antigenic peptides to major histocompatibility complex (MHC) class I molecules is a prerequisite for their recognition by cytotoxic T-cells. Prediction of MHC-binding peptides must therefore be incorporated in any predictive algorithm attempting to identify immunodominant T-cell epitopes, based on the amino acid sequence of the protein antigen. Development of predictive algorithms based on experimental binding data requires experimental testing of a very large number of peptides. A complementary approach relies on the structural conservation observed in crystallographically solved peptide-MHC complexes. By this approach, the peptide structure in the MHC groove is used as a template upon which peptide candidates are threaded, and their compatibility to bind is evaluated by statistical pairwise potentials. Our original algorithm based on this approach used the pairwise potential table of Miyazawa and Jernigan (Miyazawa S, Jernigan RL, 1996, J Mol Biol 256:623-644) and succeeded to correctly identify good binders only for MHC molecules with hydrophobic binding pockets, probably because of the high emphasis of hydrophobic interactions in this table. A recently developed pairwise potential table by Betancourt and Thirumalai (Betancourt MR, Thirumalai D, 1999, Protein Sci 8:361-369) that is based on the Miyazawa and Jernigan table describes the hydrophilic interactions more appropriately. In this paper, we demonstrate how the use of this table, together with a new definition of MHC contact residues by which only residues that contribute exclusively to sequence specific binding are included, allows the development of an improved algorithm that can be applied to a wide range of MHC class I alleles.

Keywords: immunoinformatics

Many areas of science depend on exploratory data analysis and visualization. The need to analyze large amounts of multivariate data raises the fundamental problem of dimensionality reduction: how to discover compact representations of high-dimensional data. Here, we introduce locally linear embedding (LLE), an unsupervised learning algorithm that computes low-dimensional, neighborhood-preserving embeddings of high-dimensional inputs. Unlike clustering methods for local dimensionality reduction, LLE maps its inputs into a single global coordinate system of lower dimensionality, and its optimizations do not involve local minima. By exploiting the local symmetries of linear reconstructions, LLE is able to learn the global structure of nonlinear manifolds, such as those generated by images of faces or documents of text.

Keywords: dimred

The IMGT/HLA Database is a specialist database for sequences of the human major histocompatibility (MHC) system. It includes all the HLA sequences officially recognised and named by the WHO Nomenclature Committee for Factors of the HLA System. The database provides users with online tools and facilities for the retrieval and analysis of these sequences. These include allele reports, alignment tools and a detailed database of all source cells. The online IMGT/HLA submission tool allows the submission of both new and confirmatory allele sequences directly to the WHO Nomenclature Committee for Factors of the HLA System. The latest version (release 1.4.1, November 1999) contains 1,015 HLA alleles from over 2,270 component sequences derived from the EMBL/GenBank/DDBJ databases. From its release in December 1998 until December 1999 the IMGT/HLA website received approximately 100,000 hits. The database currently focuses on the human major histocompatibility complex but will be used as a model system to provide specialist databases for the MHC sequences of other species.

Keywords: Base Sequence; Databases, Factual; Humans; Major Histocompatibility Complex; Molecular Sequence Data

The learning properties of finite-size polynomial support vector machines are analyzed in the case of realizable classification tasks. The normalization of the high-order features acts as a squeezing factor, introducing a strong anisotropy in the patterns distribution in feature space. As a function of the training set size, the corresponding generalization error presents a crossover, more or less abrupt depending on the distribution's anisotropy and on the task to be learned, between a fast-decreasing and a slowly decreasing regime. This behavior corresponds to the stepwise decrease found by Dietrich et al. [Phys. Rev. Lett. 82, 2975 (1999)] in the thermodynamic limit. The theoretical results are in excellent agreement with the numerical simulations.

Keywords: Acute, Acute Disease, Adenocarcinoma, Algorithms, Amino Acid Sequence, Animals, Artificial Intelligence, Automated, B-Lymphocytes, Bacterial Proteins, Base Pair Mismatch, Base Sequence, Bayes Theorem, Binding Sites, Biological, Bone Marrow Cells, Brachyura, Cell Compartmentation, Chemistry, Child, Chromosome Aberrations, Classification, Codon, Colonic Neoplasms, Comparative Study, Computational Biology, Computer Simulation, Computer-Assisted, DNA, Data Interpretation, Databases, Decision Trees, Diabetes Mellitus, Diagnosis, Discriminant Analysis, Discrimination Learning, Electric Conductivity, Electrophysiology, Escherichia coli Proteins, Factual, Feedback, Female, Fungal, Gastric Emptying, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Markers, Genetic Predisposition to Disease, Genomics, Hemolysins, Humans, Indians, Initiator, Ion Channels, Kinetics, Leukemia, Likelihood Functions, Lipid Bilayers, Logistic Models, Lymphocytic, Male, Markov Chains, Melanoma, Models, Molecular, Myeloid, Neoplasm, Neoplasms, Neoplastic, Neural Networks (Computer), Neurological, Nevus, Non-P.H.S., Non-U.S. Gov't, Nonlinear Dynamics, Normal Distribution, North American, Nucleic Acid Conformation, Oligonucleotide Array Sequence Analysis, Organ Specificity, Organelles, Ovarian Neoplasms, Ovary, P.H.S., Pattern Recognition, Physical, Pigmented, Predictive Value of Tests, Promoter Regions (Genetics), Protein Biosynthesis, Protein Folding, Protein Structure, Proteins, Proteome, RNA, Reproducibility of Results, Research Support, Saccharomyces cerevisiae, Secondary, Sensitivity and Specificity, Sequence Alignment, Sequence Analysis, Sex Characteristics, Skin Diseases, Skin Neoplasms, Skin Pigmentation, Software, Sound Spectrography, Statistical, Stomach Diseases, T-Lymphocytes, Thermodynamics, Transcription, Transcription Factors, Tumor Markers, Type 2, U.S. Gov't, Vertebrates, 0011102066

Keywords: Internet; Molecular Biology; Sequence Alignment, methods; Software; User-Computer Interface

Understanding how DNA binding proteins control global gene expression and chromosomal maintenance requires knowledge of the chromosomal locations at which these proteins function in vivo. We developed a microarray method that reveals the genome-wide location of DNA-bound proteins and used this method to monitor binding of gene-specific transcription activators in yeast. A combination of location and expression profiles was used to identify genes whose expression is directly controlled by Gal4 and Ste12 as cells respond to changes in carbon source and mating pheromone, respectively. The results identify pathways that are coordinately regulated by each of the two activators and reveal previously unknown functions for Gal4 and Ste12. Genome-wide location analysis will facilitate investigation of gene regulatory networks, gene function, and genome maintenance.

The organization of chromatin into higher-order structures influences chromosome function and epigenetic gene regulation. Higher-order chromatin has been proposed to be nucleated by the covalent modification of histone tails and the subsequent establishment of chromosomal subdomains by non-histone modifier factors. Here we show that human SUV39H1 and murine Suv39h1-mammalian homologues of Drosophila Su(var)3-9 and of Schizosaccharomyces pombe clr4-encode histone H3-specific methyltransferases that selectively methylate lysine 9 of the amino terminus of histone H3 in vitro. We mapped the catalytic motif to the evolutionarily conserved SET domain, which requires adjacent cysteine-rich regions to confer histone methyltransferase activity. Methylation of lysine 9 interferes with phosphorylation of serine 10, but is also influenced by pre-existing modifications in the amino terminus of H3. In vivo, deregulated SUV39H1 or disrupted Suv39h activity modulate H3 serine 10 phosphorylation in native chromatin and induce aberrant mitotic divisions. Our data reveal a functional interdependence of site-specific H3 tail modifications and suggest a dynamic mechanism for the regulation of higher-order chromatin.

Keywords: Amino Acid Sequence; Animals; Catalytic Domain; Chromatin, chemistry/metabolism; Drosophila; Hela Cells; Histone-Lysine N-Methyltransferase; Humans; Lysine, metabolism; Methylation; Methyltransferases, genetics/metabolism; Mice; Molecular Sequence Data; Phosphorylation; Protein Conformation; Protein Methyltransferases; Protein Structure, Tertiary; Recombinant Proteins, metabolism; Repressor Proteins, genetics/metabolism; Sequence Homology, Amino Acid; Serine, metabolism; Substrate Specificity

A series of microarray experiments produces observations of differential expression for thousands of genes across multiple conditions. It is often not clear whether a set of experiments are measuring fundamentally different gene expression states or are measuring similar states created through different mechanisms. It is useful, therefore, to define a core set of independent features for the expression states that allow them to be compared directly. Principal components analysis (PCA) is a statistical technique for determining the key variables in a multidimensional data set that explain the differences in the observations, and can be used to simplify the analysis and visualization of multidimensional data sets. We show that application of PCA to expression data (where the experimental conditions are the variables, and the gene expression measurements are the observations) allows us to summarize the ways in which gene responses vary under different conditions. Examination of the components also provides insight into the underlying factors that are measured in the experiments. We applied PCA to the publicly released yeast sporulation data set (Chu et al. 1998). In that work, 7 different measurements of gene expression were made over time. PCA on the time-points suggests that much of the observed variability in the experiment can be summarized in just 2 components-i.e. 2 variables capture most of the information. These components appear to represent (1) overall induction level and (2) change in induction level over time. We also examined the clusters proposed in the original paper, and show how they are manifested in principal component space. Our results are available on the internet at http:¿www.smi.stanford.edu/project/helix/PCArray .

Support vector (SV) machines are useful tools to classify populations characterized by abrupt decreases in density functions. At least for one class of Gaussian data model the SV classifier is not an optimal one according to a mean generalization error criterion. In real world problems, we have neither Gaussian populations nor data with sharp linear boundaries. Thus, the SV (maximal margin) classifiers can lose against other methods where more than a fixed number of supporting vectors contribute in determining the final weights of the classification and prediction rules. A good alternative to the linear SV machine is a specially trained and optimally stopped SLP in a transformed feature space obtained after decorrelating and scaling the multivariate data.

Keywords: Automated, Learning, Models, Neural Networks (Computer), Neurological, Pattern Recognition, 10935455

In this work we consider the problem of universal sequential probability assignment, under self-information loss, where the machine for performing the universal probability assignment is constrained to have a finite memory. Sequential probability assignment is equivalent to lossless source coding if we ignore the number of states required to convert the probability estimate into code bits. We consider both the probabilistic setting where the sequence is generated by a probabilistic source (either Bernoulli IID source or q-th order Markov source), and the deterministic setting where the sequence is a deterministic individual sequence. We also consider the case where the universal machine is deterministic, randomized, time-invariant or time-variant. We provide in most cases lower bounds and describe finite memory universal machines whose performance, in terms of the memory size, is compared to these bounds

Human breast tumours are diverse in their natural history and in their responsiveness to treatments. Variation in transcriptional programs accounts for much of the biological diversity of human cells and tumours. In each cell, signal transduction and regulatory systems transduce information from the cell's identity to its environmental status, thereby controlling the level of expression of every gene in the genome. Here we have characterized variation in gene expression patterns in a set of 65 surgical specimens of human breast tumours from 42 different individuals, using complementary DNA microarrays representing 8,102 human genes. These patterns provided a distinctive molecular portrait of each tumour. Twenty of the tumours were sampled twice, before and after a 16-week course of doxorubicin chemotherapy, and two tumours were paired with a lymph node metastasis from the same patient. Gene expression patterns in two tumour samples from the same individual were almost always more similar to each other than either was to any other sample. Sets of co-expressed genes were identified for which variation in messenger RNA levels could be related to specific features of physiological variation. The tumours could be classified into subtypes distinguished by pervasive differences in their gene expression patterns.

Keywords: breastcancer, csbcbook, csbcbook-ch3

We derive a mean-field algorithm for binary classification with gaussian processes that is based on the TAP approach originally proposed in statistical physics of disordered systems. The theory also yields an approximate leave-one-out estimator for the generalization error, which is computed with no extra computational cost. We show that from the TAP approach, it is possible to derive both a simpler "naive" mean-field theory and support vector machines (SVMs) as limiting cases. For both mean-field algorithms and support vector machines, simulation results for three small benchmark data sets are presented. They show that one may get state-of-the-art performance by using the leave-one-out estimator for model selection and the built-in leave-one-out estimators are extremely precise when compared to the exact leave-one-out estimate. The second result is taken as strong support for the internal consistency of the mean-field approach.

Keywords: Acute, Acute Disease, Adenocarcinoma, Algorithms, Amino Acid Sequence, Animals, Artificial Intelligence, Automated, B-Lymphocytes, Bacterial Proteins, Base Pair Mismatch, Base Sequence, Bayes Theorem, Binding Sites, Biological, Bone Marrow Cells, Brachyura, Cell Compartmentation, Chemistry, Child, Chromosome Aberrations, Classification, Colonic Neoplasms, Comparative Study, Computational Biology, Computer Simulation, Computer-Assisted, DNA, Data Interpretation, Databases, Decision Trees, Diabetes Mellitus, Diagnosis, Discriminant Analysis, Discrimination Learning, Electric Conductivity, Electrophysiology, Escherichia coli Proteins, Factual, Feedback, Female, Fungal, Gastric Emptying, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Markers, Genetic Predisposition to Disease, Hemolysins, Humans, Indians, Ion Channels, Kinetics, Leukemia, Likelihood Functions, Lipid Bilayers, Logistic Models, Lymphocytic, Male, Markov Chains, Melanoma, Models, Molecular, Myeloid, Neoplasm, Neoplasms, Neoplastic, Neural Networks (Computer), Neurological, Nevus, Non-P.H.S., Non-U.S. Gov't, Nonlinear Dynamics, Normal Distribution, North American, Nucleic Acid Conformation, Oligonucleotide Array Sequence Analysis, Organ Specificity, Organelles, Ovarian Neoplasms, Ovary, P.H.S., Pattern Recognition, Physical, Pigmented, Predictive Value of Tests, Promoter Regions (Genetics), Protein Folding, Protein Structure, Proteins, Proteome, RNA, Reproducibility of Results, Research Support, Saccharomyces cerevisiae, Secondary, Sensitivity and Specificity, Sequence Alignment, Sex Characteristics, Skin Diseases, Skin Neoplasms, Skin Pigmentation, Software, Sound Spectrography, Statistical, Stomach Diseases, T-Lymphocytes, Thermodynamics, Transcription, Transcription Factors, Tumor Markers, Type 2, U.S. Gov't, 11110131

Keywords: em, semi-supervised-learning, text-classification, unlabeled-data

A modular framework is proposed for modeling and understanding the relationships between molecular profile data and other domain knowledge using a combination of generative (here, graphical models) and discriminative [Support Vector Machines (SVMs)] methods. As illustration, naive Bayes models, simple graphical models, and SVMs were applied to published transcription profile data for 1,988 genes in 62 colon adenocarcinoma tissue specimens labeled as tumor or nontumor. These unsupervised and supervised learning methods identified three classes or subtypes of specimens, assigned tumor or nontumor labels to new specimens and detected six potentially mislabeled specimens. The probability parameters of the three classes were utilized to develop a novel gene relevance, ranking, and selection method. SVMs trained to discriminate nontumor from tumor specimens using only the 50-200 top-ranked genes had the same or better generalization performance than the full repertoire of 1,988 genes. Approximately 90 marker genes were pinpointed for use in understanding the basic biology of colon adenocarcinoma, defining targets for therapeutic intervention and developing diagnostic tools. These potential markers highlight the importance of tissue biology in the etiology of cancer. Comparative analysis of molecular profile data is proposed as a mechanism for predicting the physiological function of genes in instances when comparative sequence analysis proves uninformative, such as with human and yeast translationally controlled tumour protein. Graphical models and SVMs hold promise as the foundations for developing decision support systems for diagnosis, prognosis, and monitoring as well as inferring biological networks.

Keywords: biosvm

Keywords: chemoinformatics

Keywords: biosvm

The intracellular activity of the p53 tumor suppressor protein is regulated through a feedback loop involving its transcriptional target, mdm2. We present a simple mathematical model suggesting that, under certain circumstances, oscillations in p53 and Mdm2 protein levels can emerge in response to a stress signal. A delay in p53-dependent induction of Mdm2 is predicted to be required, albeit not sufficient, for this oscillatory behavior. In line with the predictions of the model, oscillations of both p53 and Mdm2 indeed occur on exposure of various cell types to ionizing radiation. Such oscillations may allow cells to repair their DNA without risking the irreversible consequences of continuous excessive p53 activation.

Keywords: csbcbook

In drug design, often enough, no structural information on a particular receptor protein is available. However, frequently a considerable number of different ligands is known together with their measured binding affinities towards a receptor under consideration. In such a situation, a set of plausible relative superpositions of different ligands, hopefully approximating their putative binding geometry, is usually the method of choice for preparing data for the subsequent application of 3D methods that analyze the similarity or diversity of the ligands. Examples are 3D-QSAR studies, pharmacophore elucidation, and receptor modeling. An aggravating fact is that ligands are usually quite flexible and a rigorous analysis has to incorporate molecular flexibility. We review the past six years of scientific publishing on molecular superposition. Our focus lies on automatic procedures to be performed on arbitrary molecular structures. Methodical aspects are our main concern here. Accordingly, plain application studies with few methodical elements are omitted in this presentation. While this review cannot mention every contribution to this actively developing field, we intend to provide pointers to the recent literature providing important contributions to computational methods for the structural alignment of molecules. Finally we provide a perspective on how superposition methods can effectively be used for the purpose of virtual database screening. In our opinion it is the ultimate goal to detect analogues in structure databases of nontrivial size in order to narrow down the search space for subsequent experiments.

Keywords: chemoinformatics

Both solid- and liquid-phase combinatorial chemistry have emerged as powerful tools for identifying pharmacologically active compounds and optimizing the biological activity of a lead compound. Complementary high-throughput in vitro assays are essential for compound evaluation. Cell-based assays that use optical endpoints permit investigation of a wide variety of functional properties of these compounds including specific intracellular biochemical pathways, protein-protein interactions, and the subcellular localization of targets. Integration of combinatorial chemistry with contemporary pharmacology now represents an important factor in drug discovery and development.

Keywords: Alzheimer Disease, Animals, Antineoplastic Agents, Biological, Bleomycin, Cell Cycle, Cell Cycle Proteins, Cell Death, Cell Line, Cell Nucleus, Cell Shape, Cell Transformation, Combinatorial Chemistry Techniques, Cultured, Drug Delivery Systems, Drug Design, Drug Evaluation, Enzyme Inhibitors, Formazans, Gene Expression, Humans, Inhibitory Concentration 50, Kinetics, Magnetic Resonance Spectroscopy, Mass, Mitochondria, Models, Molecular, Neoplasms, Neoplastic, Non-P.H.S., Non-U.S. Gov't, P.H.S., Paclitaxel, Peptide Library, Pharmaceutical Preparations, Pharmacology, Phosphoprotein Phosphatase, Preclinical, Protease Inhibitors, Protein-Tyrosine-Phosphatase, Research Support, Sensitivity and Specificity, Signal Transduction, Spectrum Analysis, Stereoisomerism, Structure-Activity Relationship, Sulfonic Acids, Tetrazolium Salts, Thiazoles, Toxicity Tests, Tumor, Tumor Cells, U.S. Gov't, cdc25 Phosphatase, 10869367

Alcoholism has a significant genetic basis, and identifying genes that confer a susceptibility to alcoholism will aid clinicians in preventing and effectively treating the disease. One commonly used technique to identify genetic risk factors for complex disorders such as alcoholism is the candidate gene approach, which directly tests the effects of genetic variants of a potentially contributing gene in an association study. These studies, which may include members of an affected family or unrelated cases and controls, can be performed relatively quickly and inexpensively and may allow identification of genes with small effects. However, the candidate gene approach is limited by how much is known of the biology of the disease being investigated. As researchers identify potential candidate genes using animal studies or linking them to DNA regions implicated through other analyses, the candidate gene approach will continue to be commonly used.

Keywords: Alcoholism; Animals; Chromosome Mapping; Disease Models, Animal; Genetic Predisposition to Disease; Genetic Variation; Humans; Mice; Mutation; Pedigree; Polymorphism, Genetic; Quantitative Trait, Heritable

Keywords: lasso

Structure-based design has emerged as a new tool in medicinal chemistry. A prerequisite for this new approach is an understanding of the principles of molecular recognition in protein-ligand complexes. If the three-dimensional structure of a given protein is known, this information can be directly exploited for the retrieval and design of new ligands. Structure-based ligand design is an iterative approach. First of all, it requires the crystal structure or a model derived from the crystal structure of a closely related homolog of the target protein, preferentially complexed with a ligand. This complex unravels the binding mode and conformation of a ligand under investigation and indicates the essential aspects determining its binding affinity. It is then used to generate new ideas about ways of improving an existing ligand or of developing new alternative bonding skeletons. Computational methods supplemented by molecular graphics are applied to assist this step of hypothesis generation. The features of the protein binding pocket can be translated into queries used for virtual computer screening of large compound libraries or to design novel ligands de novo. These initial proposals must be confirmed experimentally. Subsequently they are optimized toward higher affinity and better selectivity. The latter aspect is of utmost importance in defining and controlling the pharmacological profile of a ligand. A prerequisite to tailoring selectivity by rational design is a detailed understanding of molecular parameters determining selectivity. Taking examples from current drug development programs (HIV proteinase, t-RNA transglycosylase, thymidylate synthase, thrombin and, related serine proteinases), we describe recent advances in lead discovery via computer screening, iterative design, and understanding of selectivity discrimination.

Keywords: Animals, Chemistry, Computer Simulation, Cross-Over Studies, Crystallography, Deglutition, Deglutition Disorders, Drug Design, Endoscopy, Enzyme Inhibitors, Female, Fluoroscopy, Glossopharyngeal Nerve, HIV Protease Inhibitors, Horse Diseases, Horses, Male, Models, Molecular, Nerve Block, Non-U.S. Gov't, P.H.S., Pharmaceutical, Proteins, Quantitative Structure-Activity Relationship, Random Allocation, Research Support, Thrombin, Thymidylate Synthase, U.S. Gov't, X-Ray, 10954196

Keywords: biosvm

Keywords: herg

Keywords: csbcbook, csbcbook-mustread

G protein-coupled, seven-transmembrane segment receptors (GPCRs or 7TM receptors), with more than 1000 different members, comprise the largest superfamily of proteins in the body. Since the cloning of the first receptors more than a decade ago, extensive experimental work has uncovered multiple aspects of their function and challenged many traditional paradigms. However, it is only recently that we are beginning to gain insight into some of the most fundamental questions in the molecular function of this class of receptors. How can, for example, so many chemically diverse hormones, neurotransmitters, and other signaling molecules activate receptors believed to share a similar overall tertiary structure? What is the nature of the physical changes linking agonist binding to receptor activation and subsequent transduction of the signal to the associated G protein on the cytoplasmic side of the membrane and to other putative signaling pathways? The goal of the present review is to specifically address these questions as well as to depict the current awareness about GPCR structure-function relationships in general.

Keywords: Animals; GTP-Binding Proteins; Humans; Ligands; Models, Biological; Molecular Conformation; Receptors, Cell Surface

Whole-genome expression profiles provide a rich new data-trove for bioinformatics. Initial analyses of the profiles have included clustering and cross-referencing to 'external' information on protein structure and function. Expression profile clusters do relate to protein function, but the correlation is not perfect, with the discrepancies partially resulting from the difficulty in consistently defining function. Other attributes of proteins can also be related to expression-in particular, structure and localization-and sometimes show a clearer relationship than function.

Caspases (cysteine-containing aspartate-specific proteases) are at the core of the cell's suicide machinery. These enzymes, once activated, dismantle the cell by selectively cleaving key proteins after aspartate residues. The events culminating in caspase activation are the subject of intense study because of their role in cancer, and neurodegenerative and autoimmune disorders. Here we present a mechanistic mathematical model, formulated on the basis of newly emerging information, describing key elements of receptor-mediated and stress-induced caspase activation. We have used mass-conservation principles in conjunction with kinetic rate laws to formulate ordinary differential equations that describe the temporal evolution of caspase activation. Qualitative strategies for the prevention of caspase activation are simulated and compared with experimental data. We show that model predictions are consistent with available information. Thus, the model could aid in better understanding caspase activation and identifying therapeutic approaches promoting or retarding apoptotic cell death.

Keywords: csbcbook

Motivation: DNA microarray experiments generating thousands of gene expression measurements, are being used to gather information from tissue and cell samples regarding gene expression differences that will be useful in diagnosing disease. We have developed a new method to analyse this kind of data using support vector machines (SVMs). This analysis consists of both classification of the tissue samples, and an exploration of the data for mis-labeled or questionable tissue results. Results: We demonstrate the method in detail on samples consisting of ovarian cancer tissues, normal ovarian tissues, and other normal tissues. The dataset consists of expression experiment results for 97802 cDNAs for each tissue. As a result of computational analysis, a tissue sample is discovered and confirmed to be wrongly labeled. Upon correction of this mistake and the removal of an outlier, perfect classification of tissues is achieved, but not with high confidence. We identify and analyse a subset of genes from the ovarian dataset whose expression is highly differentiated between the types of tissues. To show robustness of the SVM method, two previously published datasets from other types of tissues or cells are analysed. The results are comparable to those previously obtained. We show that other machine learning methods also perform comparably to the SVM on many of those datasets. Availability: The SVM software is available at http://www.cs.columbia.edu/ bgrundy/svm. Contact: booch@cse.ucsc.edu

Keywords: biosvm

DNA hybridization arrays simultaneously measure the expression level for thousands of genes. These measurements provide a "snapshot" of transcription levels within the cell. A major challenge in computational biology is to uncover, from such measurements, gene/protein interactions and key biological features of cellular systems. In this paper, we propose a new framework for discovering interactions between genes based on multiple expression measurements. This framework builds on the use of Bayesian networks for representing statistical dependencies. A Bayesian network is a graph-based model of joint multivariate probability distributions that captures properties of conditional independence between variables. Such models are attractive for their ability to describe complex stochastic processes and because they provide a clear methodology for learning from (noisy) observations. We start by showing how Bayesian networks can describe interactions between genes. We then describe a method for recovering gene interactions from microarray data using tools for learning Bayesian networks. Finally, we demonstrate this method on the S. cerevisiae cell-cycle measurements of Spellman et al. (1998).

Keywords: biogm

Literature data on compounds both well- and poorly-absorbed in humans were used to build a statistical pattern recognition model of passive intestinal absorption. Robust outlier detection was utilized to analyze the well-absorbed compounds, some of which were intermingled with the poorly-absorbed compounds in the model space. Outliers were identified as being actively transported. The descriptors chosen for inclusion in the model were PSA and AlogP98, based on consideration of the physical processes involved in membrane permeability and the interrelationships and redundancies between available descriptors. These descriptors are quite straightforward for a medicinal chemist to interpret, enhancing the utility of the model. Molecular weight, while often used in passive absorption models, was shown to be superfluous, as it is already a component of both PSA and AlogP98. Extensive validation of the model on hundreds of known orally delivered drugs, "drug-like" molecules, and Pharmacopeia, Inc. compounds, which had been assayed for Caco-2 cell permeability, demonstrated a good rate of successful predictions (74-92%, depending on the dataset and exact criterion used).

Keywords: chemogenomics

An efficient node-deletion algorithm is introduced to find submatrices in expression data that have low mean squared residue scores and it is shown to perform well in finding co-regulation patterns in yeast and human. This introduces "biclustering", or simultaneous clustering of both genes and conditions, to knowledge discovery from expression data. This approach overcomes some problems associated with traditional clustering methods, by allowing automatic discovery of similarity based on a subset of attributes, simultaneous clustering of genes and conditions, and overlapped grouping that provides a better representation for genes with multiple functions or regulated by many factors.

Keywords: Algorithms; Animals; Gene Expression Profiling, methods; Humans; Multigene Family; Oligonucleotide Array Sequence Analysis, methods

Support Vector Machine (SVM), which is one kind of learning machines, was applied to predict the subcellular location of proteins from their amino acid composition. In this research, the proteins are classified into the following 12 groups: (1) chloroplast, (2) cytoplasm, (3) cytoskeleton, (4) endoplasmic reticulum, (5) extracall, (6) Golgi apparatus, (7) lysosome, (8) mitochondria, (9) nucleus, (10) peroxisome, (11) plasma membrane, and (12) vacuole, which have covered almost all the organelles and subcellular compartments in an animal or plant cell. The examination for the self-consistency and the jackknife test of the SVMs method was tested for the three sets: 2022 proteins, 2161 proteins, and 2319 proteins. As a result, the correct rate of self-consistency and jackknife test reaches 91 and 82 73 rate was tested by the three independent testing datasets containing 2240 proteins, 2513 proteins, and 2591 proteins. The correct prediction rates reach 82, 75, and 73 2591 proteins, respectively.

Keywords: biosvm

In an effort to find gene regulatory networks and clusters of genes that affect cancer susceptibility to anticancer agents, we joined a database with baseline expression levels of 7,245 genes measured by using microarrays in 60 cancer cell lines, to a database with the amounts of 5,084 anticancer agents needed to inhibit growth of those same cell lines. Comprehensive pair-wise correlations were calculated between gene expression and measures of agent susceptibility. Associations weaker than a threshold strength were removed, leaving networks of highly correlated genes and agents called relevance networks. Hypotheses for potential single-gene determinants of anticancer agent susceptibility were constructed. The effect of random chance in the large number of calculations performed was empirically determined by repeated random permutation testing; only associations stronger than those seen in multiply permuted data were used in clustering. We discuss the advantages of this methodology over alternative approaches, such as phylogenetic-type tree clustering and self-organizing maps.

Increasing numbers of methodologies are available to find functional genomic clusters in RNA expression data. We describe a technique that computes comprehensive pair-wise mutual information for all genes in such a data set. An association with a high mutual information means that one gene is non-randomly associated with another; we hypothesize this means the two are related biologically. By picking a threshold mutual information and using only associations at or above the threshold, we show how this technique was used on a public data set of 79 RNA expression measurements of 2,467 genes to construct 22 clusters, or Relevance Networks. The biological significance of each Relevance Network is explained.

Keywords: Computer Simulation; Gene Expression; Genome; Genome, Fungal; Genome, Human; Humans; Models, Genetic; Multigene Family; RNA; Saccharomyces cerevisiae

We introduce a method of functionally classifying genes by using gene expression data from DNA microarray hybridization experiments. The method is based on the theory of support vector machines (SVMs). SVMs are considered a supervised computer learning method because they exploit prior knowledge of gene function to identify unknown genes of similar function from expression data. SVMs avoid several problems associated with unsupervised clustering methods, such as hierarchical clustering and self-organizing maps. SVMs have many mathematical features that make them attractive for gene expression analysis, including their flexibility in choosing a similarity function, sparseness of solution when dealing with large data sets, the ability to handle large feature spaces, and the ability to identify outliers. We test several SVMs that use different similarity metrics, as well as some other supervised learning methods, and find that the SVMs best identify sets of genes with a common function using expression data. Finally, we use SVMs to predict functional roles for uncharacterized yeast ORFs based on their expression data.

Keywords: biosvm microarray

A few major advances have occurred in the area of physicochemical modeling of organic compounds during the past several years, spurred on by changes in the pharmaceutical industry. Recent advances include the ability to categorize and screen the overall physicochemical properties of potential drug candidates based entirely on their molecular structures and the ability to model the components that contribute to the oral absorption characteristics of potential drug candidates.

The most common human cancers are malignant neoplasms of the skin. Incidence of cutaneous melanoma is rising especially steeply, with minimal progress in non-surgical treatment of advanced disease. Despite significant effort to identify independent predictors of melanoma outcome, no accepted histopathological, molecular or immunohistochemical marker defines subsets of this neoplasm. Accordingly, though melanoma is thought to present with different 'taxonomic' forms, these are considered part of a continuous spectrum rather than discrete entities. Here we report the discovery of a subset of melanomas identified by mathematical analysis of gene expression in a series of samples. Remarkably, many genes underlying the classification of this subset are differentially regulated in invasive melanomas that form primitive tubular networks in vitro, a feature of some highly aggressive metastatic melanomas. Global transcript analysis can identify unrecognized subtypes of cutaneous melanoma and predict experimentally verifiable phenotypic characteristics that may be of importance to disease progression.

Constantly improving gene expression profiling technologies are expected to provide understanding and insight into cancer-related cellular processes. Gene expression data is also expected to significantly aid in the development of efficient cancer diagnosis and classification platforms. In this work we examine three sets of gene expression data measured across sets of tumor(s) and normal clinical samples: The first set consists of 2,000 genes, measured in 62 epithelial colon samples (Alon et al., 1999). The second consists of approximately equal to 100,000 clones, measured in 32 ovarian samples (unpublished extension of data set described in Schummer et al. (1999)). The third set consists of approximately equal to 7,100 genes, measured in 72 bone marrow and peripheral blood samples (Golub et al, 1999). We examine the use of scoring methods, measuring separation of tissue type (e.g., tumors from normals) using individual gene expression levels. These are then coupled with high-dimensional classification methods to assess the classification power of complete expression profiles. We present results of performing leave-one-out cross validation (LOOCV) experiments on the three data sets, employing nearest neighbor classifier, SVM (Cortes and Vapnik, 1995), AdaBoost (Freund and Schapire, 1997) and a novel clustering-based classification technique. As tumor samples can differ from normal samples in their cell-type composition, we also perform LOOCV experiments using appropriately modified sets of genes, attempting to eliminate the resulting bias. We demonstrate success rate of at least 90 in tumor versus normal classification, using sets of selected genes, with, as well as without, cellular-contamination-related members. These results are insensitive to the exact selection mechanism, over a certain range.

Keywords: biosvm microarray

We present a new method that we call generalized discriminant analysis (GDA) to deal with nonlinear discriminant analysis using kernel function operator. The underlying theory is close to the support vector machines (SVM) insofar as the GDA method provides a mapping of the input vectors into high-dimensional feature space. In the transformed space, linear properties make it easy to extend and generalize the classical linear discriminant analysis (LDA) to nonlinear discriminant analysis. The formulation is expressed as an eigenvalue problem resolution. Using a different kernel, one can cover a wide class of nonlinearities. For both simulated data and alternate kernels, we give classification results, as well as the shape of the decision function. The results are confirmed using real data to perform seed classification.

We describe the use of singular value decomposition in transforming genome-wide expression data from genes x arrays space to reduced diagonalized "eigengenes" x "eigenarrays" space, where the eigengenes (or eigenarrays) are unique orthonormal superpositions of the genes (or arrays). Normalizing the data by filtering out the eigengenes (and eigenarrays) that are inferred to represent noise or experimental artifacts enables meaningful comparison of the expression of different genes across different arrays in different experiments. Sorting the data according to the eigengenes and eigenarrays gives a global picture of the dynamics of gene expression, in which individual genes and arrays appear to be classified into groups of similar regulation and function, or similar cellular state and biological phenotype, respectively. After normalization and sorting, the significant eigengenes and eigenarrays can be associated with observed genome-wide effects of regulators, or with measured samples, in which these regulators are overactive or underactive, respectively.

Diffuse large B-cell lymphoma (DLBCL), the most common subtype of non-Hodgkin's lymphoma, is clinically heterogeneous: 40% of patients respond well to current therapy and have prolonged survival, whereas the remainder succumb to the disease. We proposed that this variability in natural history reflects unrecognized molecular heterogeneity in the tumours. Using DNA microarrays, we have conducted a systematic characterization of gene expression in B-cell malignancies. Here we show that there is diversity in gene expression among the tumours of DLBCL patients, apparently reflecting the variation in tumour proliferation rate, host response and differentiation state of the tumour. We identified two molecularly distinct forms of DLBCL which had gene expression patterns indicative of different stages of B-cell differentiation. One type expressed genes characteristic of germinal centre B cells ('germinal centre B-like DLBCL'); the second type expressed genes normally induced during in vitro activation of peripheral blood B cells ('activated B-like DLBCL'). Patients with germinal centre B-like DLBCL had a significantly better overall survival than those with activated B-like DLBCL. The molecular classification of tumours on the basis of gene expression can thus identify previously undetected and clinically significant subtypes of cancer.

Keywords: csbcbook

Due to the recent progress of the DNA microarray technology, a large number of gene expression profile data are being produced. How to analyze gene expression data is an important topic in computational molecular biology. Several studies have been done using the Boolean network as a model of a genetic network. This paper proposes efficient algorithms for identifying Boolean networks of bounded indegree and related biological networks, where identification of a Boolean network can be formalized as a problem of identifying many Boolean functions simultaneously. For the identification of a Boolean network, an O(mnD+1) time naive algorithm and a simple O (mnD) time algorithm are known, where n denotes the number of nodes, m denotes the number of examples, and D denotes the maximum in degree. This paper presents an improved O(momega-2nD + mnD+omega-3) time Monte-Carlo type randomized algorithm, where omega is the exponent of matrix multiplication (currently, omega < 2.376). The algorithm is obtained by combining fast matrix multiplication with the randomized fingerprint function for string matching. Although the algorithm and its analysis are simple, the result is nontrivial and the technique can be applied to several related problems.

Keywords: chemoinformatics

The Escherichia coli MG1655 genome has been completely sequenced. The annotated sequence, biochemical information, and other information were used to reconstruct the E. coli metabolic map. The stoichiometric coefficients for each metabolic enzyme in the E. coli metabolic map were assembled to construct a genome-specific stoichiometric matrix. The E. coli stoichiometric matrix was used to define the system's characteristics and the capabilities of E. coli metabolism. The effects of gene deletions in the central metabolic pathways on the ability of the in silico metabolic network to support growth were assessed, and the in silico predictions were compared with experimental observations. It was shown that based on stoichiometric and capacity constraints the in silico analysis was able to qualitatively predict the growth potential of mutant strains in 86% of the cases examined. Herein, it is demonstrated that the synthesis of in silico metabolic genotypes based on genomic, biochemical, and strain-specific information is possible, and that systems analysis methods are available to analyze and interpret the metabolic phenotype.

Keywords: Animals; Computer Communication Networks; Databases, Factual; Eukaryotic Cells; Genes; Humans; Metaphysics; Mice; Molecular Biology; Sequence Analysis, DNA; Terminology as Topic

A class of predictive densities is derived by weighting the observed samples in maximizing the log-likelihood function. This approach is effective in cases such as sample surveys or design of experiments, where the observed covariate follows a different distribution than that in the whole population. Under misspecification of the parametric model, the optimal choice of the weight function is asymptotically shown to be the ratio of the density function of the covariate in the population to that in the observations. This is the pseudo-maximum likelihood estimation of sample surveys. The optimality is defined by the expected Kullback&#x2013;Leibler loss, and the optimal weight is obtained by considering the importance sampling identity. Under correct specification of the model, however, the ordinary maximum likelihood estimate (i.e. the uniform weight) is shown to be optimal asymptotically. For moderate sample size, the situation is in between the two extreme cases, and the weight function is selected by minimizing a variant of the information criterion derived as an estimate of the expected loss. The method is also applied to a weighted version of the Bayesian predictive density. Numerical examples as well as Monte-Carlo simulations are shown for polynomial regression. A connection with the robust parametric estimation is discussed.

Keywords: domain-adaptation

To facilitate studies of the yeast proteome, we cloned 5800 open reading frames and overexpressed and purified their corresponding proteins. The proteins were printed onto slides at high spatial density to form a yeast proteome microarray and screened for their ability to interact with proteins and phospholipids. We identified many new calmodulin- and phospholipid-interacting proteins; a common potential binding motif was identified for many of the calmodulin-binding proteins. Thus, microarrays of an entire eukaryotic proteome can be prepared and screened for diverse biochemical activities. The microarrays can also be used to screen protein-drug interactions and to detect posttranslational modifications.

Keywords: Amino Acid Motifs, Amino Acid Sequence, Calmodulin, Calmodulin-Binding Proteins, Cell Membrane, Cloning, Fungal Proteins, Glucose, Liposomes, Membrane Proteins, Molecular, Molecular Sequence Data, Non-U.S. Gov't, Open Reading Frames, P.H.S., Peptide Library, Phosphatidylcholines, Phosphatidylinositols, Phospholipids, Protein Binding, Proteome, Recombinant Fusion Proteins, Research Support, Saccharomyces cerevisiae, Signal Transduction, Streptavidin, U.S. Gov't, 11474067

Using gene expression data to classify tumor types is a very promising tool in cancer diagnosis. Previous works show several pairs of tumor types can be successfully distinguished by their gene expression patterns (Golub et al. 1999, Ben-Dor et al. 2000, Alizadeh et al. 2000). However, the simultaneous classification across a heterogeneous set of tumor types has not been well studied yet. We obtained 190 samples from 14 tumor classes and generated a combined expression dataset containing 16063 genes for each of those samples. We performed multi-class classification by combining the outputs of binary classifiers. Three binary classifiers (k-nearest neighbors, weighted voting, and support vector machines) were applied in conjunction with three combination scenarios (one-vs-all, all-pairs, hierarchical partitioning). We achieved the best cross validation error rate of 18.75 support vector machine algorithm. The results demonstrate the feasibility of performing clinically useful classification from samples of multiple tumor types.

Keywords: biosvm

The posterior predictive check (PPC) is a model evaluation tool. It assigns a value (p PPC ) to the probability that the value of a given statistic computed from data arising under an analysis model is as or more extreme than the value computed from the real data themselves. If this probability is too small, the analysis model is regarded as invalid for the given statistic. Properties of the PPC for pharmacokinetic (PK) and pharmacodynamic (PD) model evaluation are examined herein for a particularly simple simulation setting: extensive sampling of a single individual's data arising from simple PK/PD and error models. To test the performance characteristics of the PPC, repeatedly, ldquorealrdquo data are simulated and for a variety of statistics, the PPC is applied to an analysis model, which may (null hypothesis) or may not (alternative hypothesis) be identical to the simulation model. Five models are used here: (PK1) mono-exponential with proportional error, (PK2) biexponential with proportional error, (PK2epsi) biexponential with additive error, (PD1) E max model with additive error under the logit transform, and (PD2) sigmoid E max model with additive error under the logit transform. Six simulation/analysis settings are studied. The first three, (PK1/PK1), (PK2/PK2), and (PD1/PD1) evaluate whether the PPC has appropriate type-I error level, whereas the second three (PK2/PK1), (PK2epsi/PK2), and (PD2/PD1) evaluate whether the PPC has adequate power. For a set of 100 data sets simulated/analyzed under each model pair according to a stipulated extensive sampling design, the p PPC is computed for a number of statistics in three different ways (each way uses a different approximation to the posterior distribution on the model parameters). We find that in general; (i) The PPC is conservative under the null in the sense that for many statistics, prob(p PPC leagr)agr for small agr. With respect to such statistics, this means that useful models will rarely be regarded incorrectly as invalid. A high correlation of a statistic with the parameter estimates obtained from the same data used to compute the statistic (a measure of statistical ldquosufficiencyrdquo) tends to identify the most conservative statistics. (ii) Power is not very great, at least for the alternative models we tested, and it is especially poor with ldquostatisticsrdquo that are in part a function of parameters as well as data. Although there is a tendency for nonsufficient statistics (as we have measured this) to have greater power, this is by no means an infallible diagnostic. (iii) No clear advantage for one or another method of approximating the posterior distribution on model parameters is found.

Mini-fingerprints (MFPs) are short binary bit string representations of molecular structure and properties, composed of few selected two-dimensional (2D) descriptors and a number of structural keys. MFPs were specifically designed to recognize compounds with similar activity. Here we report that MFPs are capable of detecting similar activities of some druglike molecules, including endothelin A antagonists and alpha(1)-adrenergic receptor ligands, the recognition of which was previously thought to depend on the use of multiple point three-dimensional (3D) pharmacophore methods. Thus, in these cases, MFPs and pharmacophore fingerprints produce similar results, although they define, in terms of their complexity, opposite ends of the spectrum of methods currently used to study molecular similarity or diversity. For each of the studied compound classes, comparison of MFP bit settings identified a consensus or signature pattern. Scaling factors can be applied to these bits in order to increase the probability of finding compounds with similar activity by virtual screening.

Keywords: Adrenergic, Angiotensin II, Cell Surface, Combinatorial Chemistry Techniques, Databases, Drug Evaluation, Endothelins, Environmental Pollutants, Factual, Information Management, Ligands, Molecular Structure, Pharmaceutical Preparations, Platelet Glycoprotein GPIIb-IIIa Complex, Preclinical, Receptors, Serine Proteinase Inhibitors, Structure-Activity Relationship, User-Computer Interface, alpha-1, 11277728

Results of systematic virtual screening calculations using a structural key-type fingerprint are reported for compounds belonging to 14 activity classes added to randomly selected synthetic molecules. For each class, a fingerprint profile was calculated to monitor the relative occupancy of fingerprint bit positions. Consensus bit patterns were determined consisting of all bits that were always set on in compounds belonging to a specific activity class. In virtual screening calculations, scale factors were applied to each consensus bit position in fingerprints of query molecules. This technique, called "fingerprint scaling", effectively increases the weight of consensus bit positions in fingerprint comparisons. Although overall prediction accuracy was satisfactory using unscaled calculations, scaling significantly increased the number of correct predictions but only slightly increased the rate of false positives. These observations suggest that fingerprint scaling is an attractive approach to increase the probability of identifying molecules with similar activity by virtual screening. It requires the availability of a series of related compounds and can be easily applied to any keyed fingerprint representation that associates bit positions with specific molecular features.

Keywords: 16S, Algae, Algorithms, Animals, Archaeal, Automation, Bacteria, Biodiversity, Chemical, Colorimetry, Computational Biology, Computer Terminals, DNA, DNA Fingerprinting, Daphnia, Databases, Ecosystem, Euryarchaeota, Factual, Fresh Water, Hazardous Substances, Humans, Information Storage and Retrieval, Methane, Models, Non-U.S. Gov't, Oxidoreductases, Perciformes, Photic Stimulation, Photometry, Polymorphism, Quantitative Structure-Activity Relationship, RNA, Research Support, Restriction Fragment Length, Ribosomal, Seasons, Soil Microbiology, Spain, Sulfur, Theoretical, Time Factors, Toxicity Tests, Water Microbiology, Water Pollutants, 11410055

Gene expression studies bridge the gap between DNA information and trait information by dissecting biochemical pathways into intermediate components between genotype and phenotype. These studies open new avenues for identifying complex disease genes and biomarkers for disease diagnosis and for assessing drug efficacy and toxicity. However, the majority of analytical methods applied to gene expression data are not efficient for biomarker identification and disease diagnosis. In this paper, we propose a general framework to incorporate feature (gene) selection into pattern recognition in the process to identify biomarkers. Using this framework, we develop three feature wrappers that search through the space of feature subsets using the classification error as measure of goodness for a particular feature subset being "wrapped around": linear discriminant analysis, logistic regression, and support vector machines. To effectively carry out this computationally intensive search process, we employ sequential forward search and sequential forward floating search algorithms. To evaluate the performance of feature selection for biomarker identification we have applied the proposed methods to three data sets. The preliminary results demonstrate that very high classification accuracy can be attained by identified composite classifiers with several biomarkers.

Keywords: biosvm

Acetylation of core histone tails plays a fundamental role in transcription regulation. In addition to acetylation, other posttranslational modifications, such as phosphorylation and methylation, occur in core histone tails. Here, we report the purification, molecular identification, and functional characterization of a histone H4-specific methyltransferase PRMT1, a protein arginine methyltransferase. PRMT1 specifically methylates arginine 3 (Arg 3) of H4 in vitro and in vivo. Methylation of Arg 3 by PRMT1 facilitates subsequent acetylation of H4 tails by p300. However, acetylation of H4 inhibits its methylation by PRMT1. Most important, a mutation in the S-adenosyl-l-methionine-binding site of PRMT1 substantially crippled its nuclear receptor coactivator activity. Our finding reveals Arg 3 of H4 as a novel methylation site by PRMT1 and indicates that Arg 3 methylation plays an important role in transcriptional regulation.

Keywords: Acetylation; Amino Acid Sequence; Animals; Arginine, metabolism; Binding Sites; Cell Nucleus, metabolism; Hela Cells; Histones, chemistry/metabolism; Humans; Hydroxamic Acids, pharmacology; Lysine, metabolism; Methylation; Methyltransferases, chemistry/genetics/isolation /&/ purification/metabolism; Molecular Sequence Data; Mutation; Oocytes; Receptors, Androgen, metabolism; Recombinant Proteins, metabolism; S-Adenosylmethionine, metabolism; Transcriptional Activation; Xenopus

In the finite-alphabet context we propose four alternatives to fixed-order Markov models to estimate a conditional distribution. They consist in working with a large class of variable-length Markov models represented by context trees, and building an estimator of the conditional distribution with a risk of the same order as the risk of the best estimator for every model simultaneously, in a conditional Kullback-Leibler sense. Such estimators can be used to model complex objects like texts written in natural language and define a notion of similarity between them. This idea is illustrated by experimental results of unsupervised text clustering

RNA and DNA strands produce ionic current signatures when driven through an alpha-hemolysin channel by an applied voltage. Here we combine this nanopore detector with a support vector machine (SVM) to analyze DNA hairpin molecules on the millisecond time scale. Measurable properties include duplex stem length, base pair mismatches, and loop length. This nanopore instrument can discriminate between individual DNA hairpins that differ by one base pair or by one nucleotide.

Keywords: Acute, Acute Disease, Adenocarcinoma, Algorithms, Amino Acid Sequence, Artificial Intelligence, Automated, B-Lymphocytes, Bacterial Proteins, Base Pair Mismatch, Base Sequence, Bayes Theorem, Binding Sites, Biological, Bone Marrow Cells, Cell Compartmentation, Chemistry, Child, Chromosome Aberrations, Comparative Study, Computational Biology, Computer Simulation, Computer-Assisted, DNA, Data Interpretation, Databases, Decision Trees, Diagnosis, Discriminant Analysis, Electric Conductivity, Electrophysiology, Escherichia coli Proteins, Factual, Female, Fungal, Gastric Emptying, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Markers, Hemolysins, Humans, Ion Channels, Kinetics, Leukemia, Lipid Bilayers, Logistic Models, Lymphocytic, Male, Markov Chains, Melanoma, Models, Molecular, Myeloid, Neoplasm, Neoplastic, Neural Networks (Computer), Nevus, Non-P.H.S., Non-U.S. Gov't, Nucleic Acid Conformation, Organ Specificity, Organelles, P.H.S., Pattern Recognition, Physical, Pigmented, Predictive Value of Tests, Promoter Regions (Genetics), Protein Folding, Protein Structure, Proteins, Proteome, RNA, Reproducibility of Results, Research Support, Saccharomyces cerevisiae, Secondary, Sensitivity and Specificity, Sequence Alignment, Sex Characteristics, Skin Diseases, Skin Neoplasms, Skin Pigmentation, Software, Statistical, Stomach Diseases, T-Lymphocytes, Thermodynamics, Transcription, Transcription Factors, Tumor Markers, U.S. Gov't, 11231558

A 2.91-billion base pair (bp) consensus sequence of the euchromatic portion of the human genome was generated by the whole-genome shotgun sequencing method. The 14.8-billion bp DNA sequence was generated over 9 months from 27,271,853 high-quality sequence reads (5.11-fold coverage of the genome) from both ends of plasmid clones made from the DNA of five individuals. Two assembly strategies-a whole-genome assembly and a regional chromosome assembly-were used, each combining sequence data from Celera and the publicly funded genome effort. The public data were shredded into 550-bp segments to create a 2.9-fold coverage of those genome regions that had been sequenced, without including biases inherent in the cloning and assembly procedure used by the publicly funded group. This brought the effective coverage in the assemblies to eightfold, reducing the number and size of gaps in the final assembly over what would be obtained with 5.11-fold coverage. The two assembly strategies yielded very similar results that largely agree with independent mapping data. The assemblies effectively cover the euchromatic regions of the human chromosomes. More than 90 bp or more, and 25 or larger. Analysis of the genome sequence revealed 26,588 protein-encoding transcripts for which there was strong corroborating evidence and an additional  12,000 computationally derived genes with mouse matches or other weak supporting evidence. Although gene-dense clusters are obvious, almost half the genes are dispersed in low G+C sequence separated by large tracts of apparently noncoding sequence. Only 1.1 with 75 segmental blocks, ranging in size up to chromosomal lengths, are abundant throughout the genome and reveal a complex evolutionary history. Comparative genomic analysis indicates vertebrate expansions of genes associated with neuronal function, with tissue-specific developmental regulation, and with the hemostasis and immune systems. DNA sequence comparisons between the consensus sequence and publicly funded genome data provided locations of 2.1 million single-nucleotide polymorphisms (SNPs). A random pair of human haploid genomes differed at a rate of 1 bp per 1250 on average, but there was marked heterogeneity in the level of polymorphism across the genome. Less than 1 task of determining which SNPs have functional consequences remains an open challenge.

Keywords: genomics bio

Microarrays can measure the expression of thousands of genes to identify changes in expression between different biological states. Methods are needed to determine the significance of these changes while accounting for the enormous number of genes. We describe a method, Significance Analysis of Microarrays (SAM), that assigns a score to each gene on the basis of change in gene expression relative to the standard deviation of repeated measurements. For genes with scores greater than an adjustable threshold, SAM uses permutations of the repeated measurements to estimate the percentage of genes identified by chance, the false discovery rate (FDR). When the transcriptional response of human cells to ionizing radiation was measured by microarrays, SAM identified 34 genes that changed at least 1.5-fold with an estimated FDR of 12%, compared with FDRs of 60 and 84% by using conventional methods of analysis. Of the 34 genes, 19 were involved in cell cycle regulation and 3 in apoptosis. Surprisingly, four nucleotide excision repair genes were induced, suggesting that this repair pathway for UV-damaged DNA might play a previously unrecognized role in repairing DNA damaged by ionizing radiation.

Keywords: csbcbook, csbcbook-ch4

Microbial catabolism of phenylpropanoid compounds plays a key role in the degradation of aromatic molecules originating from the degradation of proteins and plant constituents. In this study, the regulation of the early steps in the utilisation of 3-phenylpropionate, a phenylpropanoid compound, was investigated. Expression of the hcaA gene product, which is involved in 3-phenylpropionate catabolism in Escherichia coli, was positively regulated by HcaR, a regulatory protein similar to members of the LysR regulators family. Remarkably, the expression of hcaA in the presence of 3-phenylpropionate was sharply and transiently induced at the end of the exponential growth phase. This occurred in a rpoS-independent manner. This transient induction was also mediated by HcaR. The expression of this positive regulator is negatively autoregulated, as for other members of the LysR family. The expression of hcaR is strongly repressed in the presence of glucose. Glucose-dependent repression of hcaR expression could only be partially overcome by adding exogenous cAMP.

Circuits and their involvement in complex dynamics are described in differential terms in Part I of this work. Here, we first explain why it may be appropriate to use a logical description, either by itself or in symbiosis with the differential description. The major problem of a logical description is to find an adequate way to involve time. The procedure we adopted differs radically from the classical one by its fully asynchronous character. In Sec. II we describe our "naive" logical approach, and use it to illustrate the major laws of circuitry (namely, the involvement of positive circuits in multistationarity and of negative circuits in periodicity) and in a biological example. Already in the naive description, the major steps of the logical description are to: (i) describe a model as a set of logical equations, (ii) derive the state table from the equations, (iii) derive the graph of the sequences of states from the state table, and (iv) determine which of the possible pathways will be actually followed in terms of time delays. In the following sections we consider multivalued variables where required, the introduction of logical parameters and of logical values ascribed to the thresholds, and the concept of characteristic state of a circuit. This generalized logical description provides an image whose qualitative fit with the differential description is quite remarkable. A major interest of the generalized logical description is that it implies a limited and often quite small number of possible combinations of values of the logical parameters. The space of the logical parameters is thus cut into a limited number of boxes, each of which is characterized by a defined qualitative behavior of the system. Our analysis tells which constraints on the logical parameters must be fulfilled in order for any circuit (or combination of circuits) to be functional. Functionality of a circuit will result in multistationarity (in the case of a positive circuit) or in a cycle (in the case of a negative circuit). The last sections deal with "more about time delays" and "reverse logic," an approach that aims to proceed rationally from facts to models. (c) 2001 American Institute of Physics.

This paper presents a new scheme for training MLPs which employs a relaxation method for multi-objective optimization. The algorithm works by obtaining a reduced set of solutions, from which the one with the best generalization is selected. This approach allows balancing between the training error and norm of network weight vectors, which are the two objective functions of the multi-objective optimization problem. The method is applied to classification and regression problems and compared with Weight Decay (WD), Support Vector Machines (SVMs) and standard Backpropagation (BP). It is shown that the systematic procedure for training proposed results on good generalization neural models, and outperforms traditional methods.

We consider a new neural network for data discrimination in pattern recognition applications. We refer to this as a maximum discriminating feature (MDF) neural network. Its weights are obtained in closed-form, thereby overcoming problems associated with other nonlinear neural networks. It uses neuron activation functions that are dynamically chosen based on the application. It is theoretically shown to provide nonlinear transforms of the input data that are more general than those provided by other nonlinear multilayer perceptron neural network and support-vector machine techniques for cases involving high-dimensional (image) inputs where training data are limited and the classes are not linearly separable. We experimentally verify this on synthetic examples.

The purpose of this study was to classify breast carcinomas based on variations in gene expression patterns derived from cDNA microarrays and to correlate tumor characteristics to clinical outcome. A total of 85 cDNA microarray experiments representing 78 cancers, three fibroadenomas, and four normal breast tissues were analyzed by hierarchical clustering. As reported previously, the cancers could be classified into a basal epithelial-like group, an ERBB2-overexpressing group and a normal breast-like group based on variations in gene expression. A novel finding was that the previously characterized luminal epithelial/estrogen receptor-positive group could be divided into at least two subgroups, each with a distinctive expression profile. These subtypes proved to be reasonably robust by clustering using two different gene sets: first, a set of 456 cDNA clones previously selected to reflect intrinsic properties of the tumors and, second, a gene set that highly correlated with patient outcome. Survival analyses on a subcohort of patients with locally advanced breast cancer uniformly treated in a prospective study showed significantly different outcomes for the patients belonging to the various groups, including a poor prognosis for the basal-like subtype and a significant difference in outcome for the two estrogen receptor-positive groups.

Keywords: breastcancer, csbcbook, csbcbook-ch2

Support vector machines have been very successful in pattern recognition and function estimation problems. In this paper we introduce the use of least squares support vector machines (LS-SVM's) for the optimal control of nonlinear systems. Linear and neural full static state feedback controllers are considered. The problem is formulated in such a way that it incorporates the N-stage optimal control problem as well as a least squares support vector machine approach for mapping the state space into the action space. The solution is characterized by a set of nonlinear equations. An alternative formulation as a constrained nonlinear optimization problem in less unknowns is given, together with a method for imposing local stability in the LS-SVM control scheme. The results are discussed for support vector machines with radial basis function kernel. Advantages of LS-SVM control are that no number of hidden units has to be determined for the controller and that no centers have to be specified for the Gaussian kernels when applying Mercer's condition. The curse of dimensionality is avoided in comparison with defining a regular grid for the centers in classical radial basis function networks. This is at the expense of taking the trajectory of state variables as additional unknowns in the optimization problem, while classical neural network approaches typically lead to parametric optimization problems. In the SVM methodology the number of unknowns equals the number of training data, while in the primal space the number of unknowns can be infinite dimensional. The method is illustrated both on stabilization and tracking problems including examples on swinging up an inverted pendulum with local stabilization at the endpoint and a tracking problem for a ball and beam system.

Keywords: Acute, Acute Disease, Adenocarcinoma, Algorithms, Amino Acid Sequence, Artificial Intelligence, Automated, B-Lymphocytes, Bacterial Proteins, Base Pair Mismatch, Base Sequence, Bayes Theorem, Binding Sites, Biological, Bone Marrow Cells, Cell Compartmentation, Chemistry, Child, Chromosome Aberrations, Comparative Study, Computational Biology, Computer Simulation, Computer-Assisted, DNA, Data Interpretation, Databases, Decision Trees, Diagnosis, Discriminant Analysis, Electric Conductivity, Electrophysiology, Escherichia coli Proteins, Factual, Feedback, Female, Fungal, Gastric Emptying, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Markers, Hemolysins, Humans, Ion Channels, Kinetics, Leukemia, Lipid Bilayers, Logistic Models, Lymphocytic, Male, Markov Chains, Melanoma, Models, Molecular, Myeloid, Neoplasm, Neoplastic, Neural Networks (Computer), Nevus, Non-P.H.S., Non-U.S. Gov't, Nonlinear Dynamics, Normal Distribution, Nucleic Acid Conformation, Organ Specificity, Organelles, P.H.S., Pattern Recognition, Physical, Pigmented, Predictive Value of Tests, Promoter Regions (Genetics), Protein Folding, Protein Structure, Proteins, Proteome, RNA, Reproducibility of Results, Research Support, Saccharomyces cerevisiae, Secondary, Sensitivity and Specificity, Sequence Alignment, Sex Characteristics, Skin Diseases, Skin Neoplasms, Skin Pigmentation, Software, Statistical, Stomach Diseases, T-Lymphocytes, Thermodynamics, Transcription, Transcription Factors, Tumor Markers, U.S. Gov't, 11213211

Classification of human tumors according to their primary anatomical site of origin is fundamental for the optimal treatment of patients with cancer. Here we describe the use of large-scale RNA profiling and supervised machine learning algorithms to construct a first-generation molecular classification scheme for carcinomas of the prostate, breast, lung, ovary, colorectum, kidney, liver, pancreas, bladder/ureter, and gastroesophagus, which collectively account for [ ]70 cancer-related deaths in the United States. The classification scheme was based on identifying gene subsets whose expression typifies each cancer class, and we quantified the extent to which these genes are characteristic of a specific tumor type by accurately and confidently predicting the anatomical site of tumor origin for 90 including 9 of 12 metastatic lesions. The predictor gene subsets include those whose expression is typical of specific types of normal epithelial differentiation, as well as other genes whose expression is elevated in cancer. This study demonstrates the feasibility of predicting the tissue origin of a carcinoma in the context of multiple cancer classes.

Keywords: biosvm, breastcancer

Posttranslational modifications of histone amino termini play an important role in modulating chromatin structure and function. Lysine methylation of histones has been well documented, and recently this modification has been linked to cellular processes involving gene transcription and heterochromatin assembly. However, the existence of arginine methylation on histones has remained unclear. Recent discoveries of protein arginine methyltransferases, CARM1 and PRMT1, as transcriptional coactivators for nuclear receptors suggest that histones may be physiological targets of these enzymes as part of a poorly defined transcriptional activation pathway. Here we show by using mass spectrometry that histone H4, isolated from asynchronously growing human 293T cells, is methylated at arginine 3 (Arg-3) in vivo. In support, a novel antibody directed against histone H4 methylated at Arg-3 independently demonstrates the in vivo occurrence of this modification and reveals that H4 Arg-3 methylation is highly conserved throughout eukaryotes. Finally, we show that PRMT1 is the major, if not exclusive, H4 Arg-3 methyltransfase in human 293T cells. These findings suggest a role for arginine methylation of histones in the transcription process.

Keywords: Amino Acid Motifs; Animals; Arginine, metabolism; Cell Line; Genes, Reporter; Histones, metabolism; Humans; Immunoblotting; Methylation; Protein-Arginine N-Methyltransferases, metabolism; Recombinant Fusion Proteins, genetics/metabolism

In this article we study the generalization abilities of several classifiers of support vector machine (SVM) type using a certain class of kernels that we call universal. It is shown that the soft margin algorithms with universal kernels are consistent for a large class of classification problems including some kind of noisy tasks provided that the regularization parameter is chosen well. In particular we derive a simple sufficient condition for this parameter in the case of Gaussian RBF kernels. On the one hand our considerations are based on an investigation of an approximation property-the so-called universality-of the used kernels that ensures that all continuous functions can be approximated by certain kernel expressions. This approximation property also gives a new insight into the role of kernels in these and other algorithms. On the other hand the results are achieved by a precise study of the underlying optimization problems of the classifiers. Furthermore, we show consistency for the maximal margin classifier as well as for the soft margin SVM's in the presence of large margins. In this case it turns out that also constant regularization parameters ensure consistency for the soft margin SVM's. Finally we prove that even for simple, noise free classification problems SVM's with polynomial kernels can behave arbitrarily badly.

In response to a need for a general catalog of genome variation to address the large-scale sampling designs required by association studies, gene mapping and evolutionary biology, the National Center for Biotechnology Information (NCBI) has established the dbSNP database [S.T.Sherry, M.Ward and K. Sirotkin (1999) Genome Res., 9, 677-679]. Submissions to dbSNP will be integrated with other sources of information at NCBI such as GenBank, PubMed, LocusLink and the Human Genome Project data. The complete contents of dbSNP are available to the public at website: http://www.ncbi.nlm.nih.gov/SNP. The complete contents of dbSNP can also be downloaded in multiple formats via anonymous FTP at ftp://ncbi.nlm.nih.gov/snp/.

Keywords: Animals; Biotechnology; Databases, Factual; Genetic Variation; Humans; Information Services; Internet; National Institutes of Health (U.S.); National Library of Medicine (U.S.); Polymorphism, Single Nucleotide, genetics; United States

HLA class I expression is altered in a significant fraction of the tumor types reviewed here, reflecting either immune pressure or, simply, the accumulation of pathological changes and alterations. However, in all tumor types analyzed, a majority of the tumors express HLA class I. with a general tendency for the more severe alterations to be found in later-stage and less differentiated tumors. These results are encouraging for the development of specific immunotherapies, especially considering that (1) the relatively low sensitivity of immunohistochemical techniques might underestimate HLA expression in tumors, (2) class I expression can be induced in tumor cells as a result of local inflammation and lymphokine release, and (3) class I-negative cells would be predicted to be sensitive to Iysis by natural killer cells.

Keywords: immunoinformatics

SCF ubiquitin ligases are composed of Skp1, Cdc53, Hrt1 and one member of a large family of substrate receptors known as F-box proteins (FBPs). Here we report the identification, using sequential rounds of epitope tagging, affinity purification and mass spectrometry, of 16 Skp1 and Cdc53-associated proteins in budding yeast, including all components of SCF, 9 FBPs, Yjr033 (Rav1) and Ydr202 (Rav2). Rav1, Rav2 and Skp1 form a complex that we have named 'regulator of the (H+)-ATPase of the vacuolar and endosomal membranes' (RAVE), which associates with the V1 domain of the vacuolar membrane (H+)-ATPase (V-ATPase). V-ATPases are conserved throughout eukaryotes, and have been implicated in tumour metastasis and multidrug resistance, and here we show that RAVE promotes glucose-triggered assembly of the V-ATPase holoenzyme. Previous systematic genome-wide two-hybrid screens yielded 17 proteins that interact with Skp1 and Cdc53, only 3 of which overlap with those reported here. Thus, our results provide a distinct view of the interactions that link proteins into a comprehensive cellular network.

Keywords: Affinity, Affinity Labels, Amino Acid Sequence, Animals, Cell Cycle Proteins, Cells, Chromatography, Cloning, Comparative Study, Cullin Proteins, Cultured, Cytoplasm, DNA, DNA Damage, DNA Repair, Electrospray Ionization, Fungal, Fungal Proteins, Gene Targeting, Genetic, Glucose, Holoenzymes, Humans, Macromolecular Substances, Mass, Matrix-Assisted Laser Desorption-Ionization, Mitosis, Molecular, Molecular Sequence Data, Non-P.H.S., Non-U.S. Gov't, P.H.S., Phosphoric Monoester Hydrolases, Protein Binding, Protein Interaction Mapping, Protein Kinases, Proteome, Proteomics, Proton-Translocating ATPases, Recombinant Fusion Proteins, Research Support, Ribonucleoproteins, Ribosomes, S-Phase Kinase-Associated Proteins, Saccharomyces cerevisiae, Saccharomyces cerevisiae Proteins, Sensitivity and Specificity, Sequence Alignment, Signal Transduction, Species Specificity, Spectrometry, Spectrum Analysis, Transcription, U.S. Gov't, Vacuolar Proton-Translocating ATPases, 11283612

We describe a map of 1.42 million single nucleotide polymorphisms (SNPs) distributed throughout the human genome, providing an average density on available sequence of one SNP every 1.9 kilobases. These SNPs were primarily discovered by two projects: The SNP Consortium and the analysis of clone overlaps by the International Human Genome Sequencing Consortium. The map integrates all publicly available SNPs with described genes and other genomic features. We estimate that 60,000 SNPs fall within exon (coding and untranslated regions), and 85% of exons are within 5 kb of the nearest SNP. Nucleotide diversity varies greatly across the genome, in a manner broadly consistent with a standard population genetic model of human history. This high-density SNP map provides a public resource for defining haplotype variation across the genome, and should help to identify biomedically important genes for diagnosis and therapy.

Keywords: Chromosome Mapping; Genetic Variation; Genetics, Medical; Genetics, Population; Genome, Human; Humans; Nucleotides; Polymorphism, Single Nucleotide

We study the typical learning properties of the recently introduced soft margin classifiers (SMCs), learning realizable and unrealizable tasks, with the tools of statistical mechanics. We derive analytically the behavior of the learning curves in the regime of very large training sets. We obtain exponential and power laws for the decay of the generalization error towards the asymptotic value, depending on the task and on general characteristics of the distribution of stabilities of the patterns to be learned. The optimal learning curves of the SMCs, which give the minimal generalization error, are obtained by tuning the coefficient controlling the trade-off between the error and the regularization terms in the cost function. If the task is realizable by the SMC, the optimal performance is better than that of a hard margin support vector machine and is very close to that of a Bayesian classifier.

Orthologs are genes in different species that originate from a single gene in the last common ancestor of these species. Such genes have often retained identical biological roles in the present-day organisms. It is hence important to identify orthologs for transferring functional information between genes in different organisms with a high degree of reliability. For example, orthologs of human proteins are often functionally characterized in model organisms. Unfortunately, orthology analysis between human and e.g. invertebrates is often complex because of large numbers of paralogs within protein families. Paralogs that predate the species split, which we call out-paralogs, can easily be confused with true orthologs. Paralogs that arose after the species split, which we call in-paralogs, however, are bona fide orthologs by definition.Orthologs and in-paralogs are typically detected with phylogenetic methods, but these are slow and difficult to automate. Automatic clustering methods based on two-way best genome-wide matches on the other hand, have so far not separated in-paralogs from out-paralogs effectively.We present a fully automatic method for finding orthologs and in-paralogs from two species. Ortholog clusters are seeded with a two-way best pairwise match, after which an algorithm for adding in-paralogs is applied. The method bypasses multiple alignments and phylogenetic trees, which can be slow and error-prone steps in classical ortholog detection. Still, it robustly detects complex orthologous relationships and assigns confidence values for both orthologs and in-paralogs. The program, called INPARANOID, was tested on all completely sequenced eukaryotic genomes. To assess the quality of INPARANOID results, ortholog clusters were generated from a dataset of worm and mammalian transmembrane proteins, and were compared to clusters derived by manual tree-based ortholog detection methods. This study led to the identification with a high degree of confidence of over a dozen novel worm-mammalian ortholog assignments that were previously undetected because of shortcomings of phylogenetic methods.A WWW server that allows searching for orthologs between human and several fully sequenced genomes is installed at http://www.cgb.ki.se/inparanoid/. This is the first comprehensive resource with orthologs of all fully sequenced eukaryotic genomes. Programs and tables of orthology assignments are available from the same location.

The goal of the computer program Adjuvant! is to allow health professionals and their patients with early breast cancer to make more informed decisions about adjuvant therapy.Actuarial analysis was used to project outcomes of patients with and without adjuvant therapy based on estimates of prognosis largely derived from Surveillance, Epidemiology, and End-Results data and estimates of the efficacy of adjuvant therapy based on the 1998 overviews of randomized trials of adjuvant therapy. These estimates can be refined using the Prognostic Factor Impact Calculator, which uses a Bayesian method to make adjustments based on relative risks conferred and prevalence of positive test results.From the entries of patient information (age, menopausal status, comorbidity estimate) and tumor staging and characteristics (tumor size, number of positive axillary nodes, estrogen receptor status), baseline prognostic estimates are made. Estimates for the efficacy of endocrine therapy (5 years of tamoxifen) and of polychemotherapy (cyclophosphamide/methotrexate/fluorouracil-like regimens, or anthracycline-based therapy, or therapy based on both an anthracycline and a taxane) can then be used to project outcomes presented in both numerical and graphical formats. Outcomes for overall survival and disease-free survival and the improvement seen in clinical trials, are reasonably modeled by Adjuvant!, although an ideal validation for all patient subsets with all treatment options is not possible. Additional speculative estimates of years of remaining life expectancy and long-term survival curves can also be produced. Help files supply general information about breast cancer. The program's Internet links supply national treatment guidelines, cooperative group trial options, and other related information.The computer program Adjuvant! can play practical and educational roles in clinical settings.

Keywords: Actuarial Analysis; Breast Neoplasms, mortality/therapy; Chemotherapy, Adjuvant; Decision Making; Female; Humans; Prognosis; Software; Survival Analysis

The optimal treatment of patients with cancer depends on establishing accurate diagnoses by using a complex combination of clinical and histopathological data. In some instances, this task is difficult or impossible because of atypical clinical presentation or histopathology. To determine whether the diagnosis of multiple common adult malignancies could be achieved purely by molecular classification, we subjected 218 tumor samples, spanning 14 common tumor types, and 90 normal tissue samples to oligonucleotide microarray gene expression analysis. The expression levels of 16,063 genes and expressed sequence tags were used to evaluate the accuracy of a multiclass classifier based on a support vector machine algorithm. Overall classification accuracy was 78 Poorly differentiated cancers resulted in low-confidence predictions and could not be accurately classified according to their tissue of origin, indicating that they are molecularly distinct entities with dramatically different gene expression patterns compared with their well differentiated counterparts. Taken together, these results demonstrate the feasibility of accurate, multiclass molecular cancer classification and suggest a strategy for future clinical implementation of molecular cancer diagnostics.

Keywords: biosvm microarray

Identification of components present in biological complexes requires their purification to near homogeneity. Methods of purification vary from protein to protein, making it impossible to design a general purification strategy valid for all cases. We have developed the tandem affinity purification (TAP) method as a tool that allows rapid purification under native conditions of complexes, even when expressed at their natural level. Prior knowledge of complex composition or function is not required. The TAP method requires fusion of the TAP tag, either N- or C-terminally, to the target protein of interest. Starting from a relatively small number of cells, active macromolecular complexes can be isolated and used for multiple applications. Variations of the method to specifically purify complexes containing two given components or to subtract undesired complexes can easily be implemented. The TAP method was initially developed in yeast but can be successfully adapted to various organisms. Its simplicity, high yield, and wide applicability make the TAP method a very useful procedure for protein purification and proteome exploration.

Keywords: Bacterial Proteins; Blotting, Western; DNA, Bacterial; Fungal Proteins; Genetic Vectors; Methods; Mutation; Polymerase Chain Reaction; Proteins; Proteome; Ribonucleases; Ribonucleoproteins; Saccharomyces cerevisiae; Saccharomyces cerevisiae Proteins; Staphylococcus aureus

Keywords: biosvm

Keywords: biosvm

The assembly of higher order chromatin structures has been linked to the covalent modifications of histone tails. We provide in vivo evidence that lysine 9 of histone H3 (H3 Lys9) is preferentially methylated by the Clr4 protein at heterochromatin-associated regions in fission yeast. Both the conserved chromo- and SET domains of Clr4 are required for H3 Lys9 methylation in vivo. Localization of Swi6, a homolog of Drosophila HP1, to heterochomatic regions is dependent on H3 Lys9 methylation. Moreover, an H3-specific deacetylase Clr3 and a beta-propeller domain protein Rik1 are required for H3 Lys9 methylation by Clr4 and Swi6 localization. These data define a conserved pathway wherein sequential histone modifications establish a "histone code" essential for the epigenetic inheritance of heterochromatin assembly.

Keywords: Acetylation; Cell Cycle Proteins, chemistry/genetics/metabolism; Centromere, metabolism; Chromosomes, Fungal, metabolism; Fungal Proteins, genetics/metabolism; Gene Silencing; Genes, Fungal; Heterochromatin, metabolism; Histone Deacetylases, genetics/metabolism; Histone-Lysine N-Methyltransferase; Histones, chemistry/metabolism; Lysine, metabolism; Methylation; Methyltransferases, chemistry/genetics/metabolism; Mutation; Protein Methyltransferases; Protein Structure, Tertiary; Recombinant Proteins, chemistry/metabolism; Saccharomyces cerevisiae Proteins; Schizosaccharomyces pombe Proteins; Schizosaccharomyces, genetics/metabolism; Transcription Factors, metabolism

Molecular portraits, such as mRNA expression or DNA methylation patterns, have been shown to be strongly correlated with phenotypical parameters. These molecular patterns can be revealed routinely on a genomic scale. However, class prediction based on these patterns is an under-determined problem, due to the extreme high dimensionality of the data compared to the usually small number of available samples. This makes a reduction of the data dimensionality necessary. Here we demonstrate how phenotypic classes can be predicted by combining feature selection and discriminant analysis. By comparing several feature selection methods we show that the right dimension reduction strategy is of crucial importance for the classification performance. The techniques are demonstrated by methylation pattern based discrimination between acute lymphoblastic leukemia and acute myeloid leukemia. Contact: Fabian.Model@epigenomics.com

Keywords: biosvm

EEG spike and wave (SW) activity has been described through a non-parametric stochastic model estimated by the Nadaraya-Watson (NW) method. In this paper the performance of the NW, the local linear polynomial regression and support vector machines (SVM) methods were compared. The noise-free realizations obtained by the NW and SVM methods reproduced SW better than as reported in previous works. The tuning parameters had to be estimated manually. Adding dynamical noise, only the NW method was capable of generating SW similar to training data. The standard deviation of the dynamical noise was estimated by means of the correlation dimension.

Keywords: Acute, Acute Disease, Adenocarcinoma, Algorithms, Amino Acid Sequence, Animals, Artificial Intelligence, Automated, B-Lymphocytes, Bacterial Proteins, Base Pair Mismatch, Base Sequence, Bayes Theorem, Binding Sites, Biological, Bone Marrow Cells, Brachyura, Cell Compartmentation, Chemistry, Child, Chromosome Aberrations, Classification, Codon, Colonic Neoplasms, Comparative Study, Computational Biology, Computer Simulation, Computer-Assisted, DNA, Data Interpretation, Databases, Decision Trees, Diabetes Mellitus, Diagnosis, Discriminant Analysis, Discrimination Learning, Electric Conductivity, Electroencephalography, Electrophysiology, Epilepsy, Escherichia coli Proteins, Factual, Feedback, Female, Fungal, Gastric Emptying, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Markers, Genetic Predisposition to Disease, Genomics, Hemolysins, Humans, Indians, Information Storage and Retrieval, Initiator, Ion Channels, Kinetics, Leukemia, Likelihood Functions, Linear Models, Lipid Bilayers, Logistic Models, Lymphocytic, MEDLINE, Male, Markov Chains, Melanoma, Models, Molecular, Myeloid, Neoplasm, Neoplasms, Neoplastic, Neural Networks (Computer), Neurological, Nevus, Non-P.H.S., Non-U.S. Gov't, Nonlinear Dynamics, Normal Distribution, North American, Nucleic Acid Conformation, Oligonucleotide Array Sequence Analysis, Organ Specificity, Organelles, Ovarian Neoplasms, Ovary, P.H.S., Pattern Recognition, Physical, Pigmented, Predictive Value of Tests, Promoter Regions (Genetics), Protein Biosynthesis, Protein Folding, Protein Structure, Proteins, Proteome, RNA, Reproducibility of Results, Research Support, Saccharomyces cerevisiae, Secondary, Sensitivity and Specificity, Sequence Alignment, Sequence Analysis, Sex Characteristics, Skin Diseases, Skin Neoplasms, Skin Pigmentation, Software, Sound Spectrography, Statistical, Stochastic Processes, Stomach Diseases, T-Lymphocytes, Thermodynamics, Transcription, Transcription Factors, Tumor Markers, Type 2, U.S. Gov't, Vertebrates, 11058693

We investigated the inhibition of cell-cycle progression and replication and the induction of the transcriptional response in diploid budding yeast populations exposed to two different doses of gamma-rays resulting in 15 and 85% survival respectively. We studied the kinetics of the cellular response to ionizing treatment during the period required for all of the surviving cells to achieve at least one cell division. The length of these periods increased with the dose. Irradiated populations arrested as large-budded cells containing partially replicated chromosomes. The extent of the S-phase was proportional to the amount of damage and lasted 3 or 7h depending on the irradiation dose. In parallel to the division study, we carried out a kinetic analysis of the expression of 126 selected genes by use of dedicated microarrays. About 26 genes were induced by irradiation and displayed various pattern of expression. Interestingly, 10 repair genes (RAD51, RAD54, CDC8, MSH2, RFA2, RFA3, UBC5, SRS2, SPO12 and TOP1), involved in recombination and DNA synthesis, display similar regulation of expression in the two irradiated populations. Their pattern of expression were confirmed by Northern analysis. At the two doses, the expression of this group of genes closely followed the extended replication period, and their expression resumed when replication restarted. These results suggest that the damage-induced response and DNA synthesis are closely regulated during repair. The analysis of the promoter regions indicates a high occurrence of the three MCB, HAP and UASH regulatory boxes in the promoters of this group of genes. The association of the three boxes could confer an irradiation-replication specific regulation.

High-throughput synthesis and screening technologies have enhanced the impact of computational chemistry on the drug discovery process. From the design of targeted, drug-like libraries to 'virtual' optimization of potency, selectivity and ADME/Tox properties, computational chemists are able to efficiently manage costly resources and dramatically shorten drug discovery cycle times. This review will describe some of the successful strategies and applications of state-of-the-art algorithms to enhance drug discovery, as well as key points in the drug discovery process where computational methods can have, and have had, greatest impact.

Keywords: chemoinformatics

Activation of gene transcription involves chromatin remodeling by coactivator proteins that are recruited by DNA-bound transcription factors. Local modification of chromatin structure at specific gene promoters by ATP-dependent processes and by posttranslational modifications of histone N-terminal tails provides access to RNA polymerase II and its accompanying transcription initiation complex. While the roles of lysine acetylation, serine phosphorylation, and lysine methylation of histones in chromatin remodeling are beginning to emerge, low levels of arginine methylation of histones have only recently been documented, and its physiological role is unknown. The coactivator CARM1 methylates histone H3 at Arg17 and Arg26 in vitro and cooperates synergistically with p160-type coactivators (e.g., GRIP1, SRC-1, ACTR) and coactivators with histone acetyltransferase activity (e.g., p300, CBP) to enhance gene activation by steroid and nuclear hormone receptors (NR) in transient transfection assays. In the current study, CARM1 cooperated with GRIP1 to enhance steroid hormone-dependent activation of stably integrated mouse mammary tumor virus (MMTV) promoters, and this coactivator function required the methyltransferase activity of CARM1. Chromatin immunoprecipitation assays and immunofluorescence studies indicated that CARM1 and the CARM1-methylated form of histone H3 specifically associated with a large tandem array of MMTV promoters in a hormone-dependent manner. Thus, arginine-specific histone methylation by CARM1 is an important part of the transcriptional activation process.

Keywords: Acetylation; Arginine, metabolism; Fluorescent Antibody Technique; Histones, chemistry/metabolism; Hormones, physiology; Lysine, metabolism; Mammary Tumor Virus, Mouse, genetics; Methylation; Phosphorylation; Precipitin Tests; Promoter Regions, Genetic; Protein-Arginine N-Methyltransferases, physiology; Serine, metabolism; Steroids, physiology

Experimental and computational approaches to estimate solubility and permeability in discovery and development settings are described. In the discovery setting 'the rule of 5' predicts that poor absorption or permeation is more likely when there are more than 5 H-bond donors, 10 H-bond acceptors, the molecular weight (MWT) is greater than 500 and the calculated Log P (CLogP) is greater than 5 (or MlogP > 4.15). Computational methodology for the rule-based Moriguchi Log P (MLogP) calculation is described. Turbidimetric solubility measurement is described and applied to known drugs. High throughput screening (HTS) leads tend to have higher MWT and Log P and lower turbidimetric solubility than leads in the pre-HTS era. In the development setting, solubility calculations focus on exact value prediction and are difficult because of polymorphism. Recent work on linear free energy relationships and Log P approaches are critically reviewed. Useful predictions are possible in closely related analog series when coupled with experimental thermodynamic solubility measurements.

Keywords: chemoinformatics

We study the recognition of surfaces made from different materials such as concrete, rug, marble, or leather on the basis of their textural appearance. Such natural textures arise from spatial variation of two surface attributes: (1) reflectance and (2) surface normal. In this paper, we provide a unified model to address both these aspects of natural texture. The main idea is to construct a vocabulary of prototype tiny surface patches with associated local geometric and photometric properties. We call these 3D textons. Examples might be ridges, grooves, spots or stripes or combinations thereof. Associated with each texton is an appearance vector, which characterizes the local irradiance distribution, represented as a set of linear Gaussian derivative filter outputs, under different lighting and viewing conditions. Given a large collection of images of different materials, a clustering approach is used to acquire a small (on the order of 100) 3D texton vocabulary. Given a few (1 to 4) images of any material, it can be characterized using these textons. We demonstrate the application of this representation for recognition of the material viewed under novel lighting and viewing conditions. We also illustrate how the 3D texton model can be used to predict the appearance of materials under novel conditions.

MELTING computes the enthalpy and entropy of an oligonucleotide duplex helix-coil transition, and then its melting temperature. The program uses the method of nearest-neighbours. The set of thermodynamic parameters can be easily customized. The program provides several correction methods for the concentration of salt. MELTING is a free program, available at no cost and open-source. Perl scripts are provided to show how MELTING can be used to construct more ambitious programs. AVAILABILITY: MELTING is available for several platforms (http://www.pasteur.fr/recherche/unites/neubiomol/meltinghome.html) and is accessible via a www server (http://bioweb.pasteur.fr/seqanal/interfaces/melting.html). CONTACT: nl223@cus.cam.ac.uk

Keywords: conditional-random-field

We examine experimental design issues arising with gene expression microarray technology. Microarray experiments have multiple sources of variation, and experimental plans should ensure that effects of interest are not confounded with ancillary effects. A commonly used design is shown to violate this principle and to be generally inefficient. We explore the connection between microarray designs and classical block design and use a family of ANOVA models as a guide to choosing a design. We combine principles of good design and A-optimality to give a general set of recommendations for design with microarrays. These recommendations are illustrated in detail for one kind of experimental objective, where we also give the results of a computer search for good designs.

Tissue microarray (TMA) technology allows rapid visualization of molecular targets in thousands of tissue specimens at a time, either at the DNA, RNA or protein level. The technique facilitates rapid translation of molecular discoveries to clinical applications. By revealing the cellular localization, prevalence and clinical significance of candidate genes, TMAs are ideally suitable for genomics-based diagnostic and drug target discovery. TMAs have a number of advantages compared with conventional techniques. The speed of molecular analyses is increased by more than 100-fold, precious tissues are not destroyed and a very large number of molecular targets can be analyzed from consecutive TMA sections. The ability to study archival tissue specimens is an important advantage as such specimens are usually not applicable in other high-throughput genomic and proteomic surveys. Construction and analysis of TMAs can be automated, increasing the throughput even further. Most of the applications of the TMA technology have come from the field of cancer research. Examples include analysis of the frequency of molecular alterations in large tumor materials, exploration of tumor progression, identification of predictive or prognostic factors and validation of newly discovered genes as diagnostic and therapeutic targets.

Keywords: Animals; Genetic Techniques; Humans; In Situ Hybridization, methods; Neoplasms, metabolism/pathology; Oligonucleotide Array Sequence Analysis; Tissue Distribution

The human genome holds an extraordinary trove of information about human development, physiology, medicine and evolution. Here we report the results of an international collaboration to produce and make freely available a draft sequence of the human genome. We also present an initial analysis of the data, describing some of the insights that can be gleaned from the sequence.

Keywords: genomics bio

Motivation: Subcellular localization is a key functional characteristic of proteins. A fully automatic and reliable prediction system for protein subcellular localization is needed, especially for the analysis of large-scale genome sequences. Results: In this paper, Support Vector Machine has been introduced to predict the subcellular localization of proteins from their amino acid compositions. The total prediction accuracies reach 91.4 in prokaryotic organisms and 79.4 organisms. Predictions by our approach are robust to errors in the protein N-terminal sequences. This new approach provides superior prediction performance compared with existing algorithms based on amino acid composition and can be a complementary method to other existing methods based on sorting signals. Availability: A web server implementing the prediction method is available at http://www.bioinfo.tsinghua.edu.cn/SubLoc/. Contact: sunzhr@mail.tsinghua.edu.cn; huasj00@mails.tsinghua.edu.cn Supplementary information: Supplementary material is available at http://www.bioinfo.tsinghua.edu.cn/SubLoc

Keywords: biosvm

Adenomatous polyps in the colon are believed to be the precursor to colorectal carcinoma, the second leading cause of cancer deaths in United States. In this paper, we propose a new method for computer-aided detection of polyps in computed tomography (CT) colonography (virtual colonoscopy), a technique in which polyps are imaged along the wall of the air-inflated, cleansed colon with X-ray CT. Initial work with computer aided detection has shown high sensitivity, but at a cost of too many false positives. We present a statistical approach that uses support vector machines to distinguish the differentiating characteristics of polyps and healthy tissue, and uses this information for the classification of the new cases. One of the main contributions of the paper is the new three-dimensional pattern processing approach, called random orthogonal shape sections method, which combines the information from many random images to generate reliable signatures of shape. The input to the proposed system is a collection of volume data from candidate polyps obtained by a high-sensitivity, low-specificity system that we developed previously. The results of our ten-fold cross-validation experiments show that, on the average, the system increases the specificity from 0.19 (0.35) to 0.69 (0.74) at a sensitivity level of 1.0 (0.95).

The global gene expression profiles for 67 human lung tumors representing 56 patients were examined by using 24,000-element cDNA microarrays. Subdivision of the tumors based on gene expression patterns faithfully recapitulated morphological classification of the tumors into squamous, large cell, small cell, and adenocarcinoma. The gene expression patterns made possible the subclassification of adenocarcinoma into subgroups that correlated with the degree of tumor differentiation as well as patient survival. Gene expression analysis thus promises to extend and refine standard pathologic analysis.

We used the support vector machines (SVM) in a classification approach to `beat the market'. Given the fundamental accounting and price information of stocks trading on the Australian Stock Exchange, we attempt to use SVM to identify stocks that are likely to outperform the market by having exceptional returns. The equally weighted portfolio formed by the stocks selected by SVM has a total return of 208 a five years period, significantly outperformed the benchmark of 71 whereby the output of SVM is interpreted as a probability measure and ranked, such that the stocks selected can be fixed to the top 25%

Double-stranded RNA (dsRNA) induces sequence-specific posttranscriptional gene silencing in many organisms by a process known as RNA interference (RNAi). Using a Drosophila in vitro system, we demonstrate that 21- and 22-nt RNA fragments are the sequence-specific mediators of RNAi. The short interfering RNAs (siRNAs) are generated by an RNase III-like processing reaction from long dsRNA. Chemically synthesized siRNA duplexes with overhanging 3' ends mediate efficient target RNA cleavage in the lysate, and the cleavage site is located near the center of the region spanned by the guiding siRNA. Furthermore, we provide evidence that the direction of dsRNA processing determines whether sense or antisense target RNA can be cleaved by the siRNA-protein complex.

Keywords: sirna

We analyze the discriminatory power of k-nearest neighbors, logistic regression, artificial neural networks (ANNs), decision tress, and support vector machines (SVMs) on the task of classifying pigmented skin lesions as common nevi, dysplastic nevi, or melanoma. Three different classification tasks were used as benchmarks: the dichotomous problem of distinguishing common nevi from dysplastic nevi and melanoma, the dichotomous problem of distinguishing melanoma from common and dysplastic nevi, and the trichotomous problem of correctly distinguishing all three classes. Using ROC analysis to measure the discriminatory power of the methods shows that excellent results for specific classification problems in the domain of pigmented skin lesions can be achieved with machine-learning methods. On both dichotomous and trichotomous tasks, logistic regression, ANNs, and SVMs performed on about the same level, with k-nearest neighbors and decision trees performing worse.

Keywords: Algorithms, Amino Acid Sequence, Artificial Intelligence, Biological, Cell Compartmentation, Comparative Study, Computer Simulation, Computer-Assisted, Decision Trees, Diagnosis, Discriminant Analysis, Humans, Logistic Models, Melanoma, Models, Neural Networks (Computer), Nevus, Non-U.S. Gov't, Organelles, P.H.S., Pigmented, Predictive Value of Tests, Proteins, Reproducibility of Results, Research Support, Skin Diseases, Skin Neoplasms, Skin Pigmentation, U.S. Gov't, 11376540

Motivation: Protein fold recognition is an important approach to structure discovery without relying on sequence similarity. We study this approach with new multi-class classification methods and examined many issues important for a practical recognition system. Results: Most current discriminative methods for protein fold prediction use the one-against-others method, which has the well-known ?False Positives? problem. We investigated two new methods: the unique one-against-others and the all-against-all methods. Both improve prediction accuracy by 14?110 SCOP folds. We used the Support Vector Machine (SVM) and the Neural Network (NN) learning methods as base classifiers. SVMs converges fast and leads to high accuracy. When scores of multiple parameter datasets are combined, majority voting reduces noise and increases recognition accuracy. We examined many issues involved with large number of classes, including dependencies of prediction accuracy on the number of folds and on the number of representatives in a fold. Overall, recognition systems achieve 56 accuracy on a protein test dataset, where most of the proteins have below 25 information: The protein parameter datasets used in this paper are available online (http://www.nersc.gov/ cding/protein).

Keywords: biosvm

A method for unsupervised segmentation of color-texture regions in images and video is presented. This method, which we refer to as JSEG, consists of two independent steps: color quantization and spatial segmentation. In the first step, colors in the image are quantized to several representative classes that can be used to differentiate regions in the image. The image pixels are then replaced by their corresponding color class labels, thus forming a class-map of the image. The focus of this work is on spatial segmentation, where a criterion for "good" segmentation using the class-map is proposed. Applying the criterion to local windows in the class-map results in the "J-image," in which high and low values correspond to possible boundaries and interiors of color-texture regions. A region growing method is then used to segment the image based on the multiscale J-images. A similar approach is applied to video sequences. An additional region tracking scheme is embedded into the region growing process to achieve consistent segmentation and tracking results, even for scenes with nonrigid object motion. Experiments show the robustness of the JSEG algorithm on real images and video.

Keywords: Aged; Antineoplastic Agents, therapeutic use; Breast Neoplasms, drug therapy/mortality; Chemotherapy, Adjuvant; Drug Therapy, Combination; Female; Follow-Up Studies; Humans; Middle Aged; Prognosis; Quality-Adjusted Life Years; Randomized Controlled Trials as Topic; Survival Analysis; Tamoxifen, therapeutic use

Transcription profiling experiments permit the expression levels of many genes to be measured simultaneously. Given profiling data from two types of samples, genes that most distinguish the samples (marker genes) are good candidates for subsequent in-depth experimental studies and developing decision support systems for diagnosis, prognosis, and monitoring. This work proposes a mixture of feature relevance experts as a method for identifying marker genes and illustrates the idea using published data from samples labeled as acute lymphoblastic and myeloid leukemia (ALL, AML). A feature relevance expert implements an algorithm that calculates how well a gene distinguishes samples, reorders genes according to this relevance measure, and uses a supervised learning method [here, support vector machines (SVMs)] to determine the generalization performances of different nested gene subsets. The mixture of three feature relevance experts examined implement two existing and one novel feature relevance measures. For each expert, a gene subset consisting of the top 50 genes distinguished ALL from AML samples as completely as all 7,070 genes. The 125 genes at the union of the top 50s are plausible markers for a prototype decision support system. Chromosomal aberration and other data support the prediction that the three genes at the intersection of the top 50s, cystatin C, azurocidin, and adipsin, are good targets for investigating the basic biology of ALL/AML. The same data were employed to identify markers that distinguish samples based on their labels of T cell/B cell, peripheral blood/bone marrow, and male/female. Selenoprotein W may discriminate T cells from B cells. Results from analysis of transcription profiling data from tumor/nontumor colon adenocarcinoma samples support the general utility of the aforementioned approach. Theoretical issues such as choosing SVM kernels and their parameters, training and evaluating feature relevance experts, and the impact of potentially mislabeled samples on marker identification (feature selection) are discussed.

Keywords: biosvm

Currently there is no successful computational approach for identification of genes encoding novel functional RNAs (fRNAs) in genomic sequences. We have developed a machine learning approach using neural networks and support vector machines to extract common features among known RNAs for prediction of new RNA genes in the unannotated regions of prokaryotic and archaeal genomes. The Escherichia coli genome was used for development, but we have applied this method to several other bacterial and archaeal genomes. Networks based on nucleotide composition were 80-90 for bacteria and 90-99 achieved a significant improvement in accuracy by combining these predictions with those obtained using a second set of parameters consisting of known RNA sequence motifs and the calculated free energy of folding. Several known fRNAs not included in the training datasets were identified as well as several hundred predicted novel RNAs. These studies indicate that there are many unidentified RNAs in simple genomes that can be predicted computationally as a precursor to experimental study. Public access to our RNA gene predictions and an interface for user predictions is available via the web.

Keywords: biosvm

Short interfering RNAs (siRNAs) are double-stranded RNAs of approximately 21-25 nucleotides that have been shown to function as key intermediaries in triggering sequence-specific RNA degradation during posttranscriptional gene silencing in plants and RNA interference in invertebrates. siRNAs have a characteristic structure, with 5'-phosphate/3'-hydroxyl ends and a 2-base 3' overhang on each strand of the duplex. In this study, we present data that synthetic siRNAs can induce gene-specific inhibition of expression in Caenorhabditis elegans and in cell lines from humans and mice. In each case, the interference by siRNAs was superior to the inhibition of gene expression mediated by single-stranded antisense oligonucleotides. The siRNAs seem to avoid the well documented nonspecific effects triggered by longer double-stranded RNAs in mammalian cells. These observations may open a path toward the use of siRNAs as a reverse genetic and therapeutic tool in mammalian cells.

Keywords: sirna

Background We apply a new machine learning method, the so-called Support Vector Machine method, to predict the protein structural class. Support Vector Machine method is performed based on the database derived from SCOP, in which protein domains are classified based on known structures and the evolutionary relationships and the principles that govern their 3-D structure. Results High rates of both self-consistency and jackknife tests are obtained. The good results indicate that the structural class of a protein is considerably correlated with its amino acid composition. Conclusions It is expected that the Support Vector Machine method and the elegant component-coupled method, also named as the covariant discrimination algorithm, if complemented with each other, can provide a powerful computational tool for predicting the structural classes of proteins.

Keywords: biosvm

Histone methylation is known to be associated with both transcriptionally active and repressive chromatin states. Recent studies have identified SET domain-containing proteins such as SUV39H1 and Clr4 as mediators of H3 lysine 9 (Lys9) methylation and heterochromatin formation. Interestingly, H3 Lys9 methylation is not observed from bulk histones isolated from asynchronous populations of Saccharomyces cerevisiae or Tetrahymena thermophila. In contrast, H3 lysine 4 (Lys4) methylation is a predominant modification in these smaller eukaryotes. To identify the responsible methyltransferase(s) and to gain insight into the function of H3 Lys4 methylation, we have developed a histone H3 Lys4 methyl-specific antiserum. With this antiserum, we show that deletion of SET1, but not of other putative SET domain-containing genes, in S. cerevisiae, results in the complete abolishment of H3 Lys4 methylation in vivo. Furthermore, loss of H3 Lys4 methylation in a set1 Delta strain can be rescued by SET1. Analysis of histone H3 mutations at Lys4 revealed a slow-growth defect similar to a set1 Delta strain. Chromatin immunoprecipitation assays show that H3 Lys4 methylation is present at the rDNA locus and that Set1-mediated H3 Lys4 methylation is required for repression of RNA polymerase II transcription within rDNA. Taken together, these data suggest that Set1-mediated H3 Lys4 methylation is required for normal cell growth and transcriptional silencing.

Keywords: Animals; Antibody Formation; Blotting, Western; Cell Division; DNA Primers, chemistry; DNA, Bacterial, genetics; DNA, Ribosomal, genetics; DNA-Binding Proteins, metabolism; Fungal Proteins, metabolism; Gene Silencing; Genetic Vectors; Heterochromatin, chemistry/metabolism; Histone-Lysine N-Methyltransferase; Histones, metabolism; Lysine, metabolism; Methylation; Methyltransferases, genetics/metabolism; Mutation; Nucleosomes, chemistry/metabolism; Polymerase Chain Reaction; Precipitin Tests; Protein Methyltransferases; RNA Polymerase III, metabolism; Rabbits; Saccharomyces cerevisiae Proteins; Saccharomyces cerevisiae, genetics; Transcription Factors, metabolism

Random forests are a combination of tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. The generalization error for forests converges a.s. to a limit as the number of trees in the forest becomes large. The generalization error of a forest of tree classifiers depends on the strength of the individual trees in the forest and the correlation between them. Using a random selection of features to split each node yields error rates that compare favorably to Adaboost (Y. Freund & R. Schapire, Machine Learning: Proceedings of the Thirteenth International conference, ***, 148–156), but are more robust with respect to noise. Internal estimates monitor error, strength, and correlation and these are used to show the response to increasing the number of features used in the splitting. Internal estimates are also used to measure variable importance. These ideas are also applicable to regression.

Keywords: PUlearning

Microarray analysis has become a widely used tool for the generation of gene expression data on a genomic scale. Although many significant results have been derived from microarray studies, one limitation has been the lack of standards for presenting and exchanging such data. Here we present a proposal, the Minimum Information About a Microarray Experiment (MIAME), that describes the minimum information required to ensure that microarray data can be easily interpreted and that results derived from its analysis can be independently verified. The ultimate goal of this work is to establish a standard for recording and reporting microarray-based gene expression data, which will in turn facilitate the establishment of databases and public repositories and enable the development of data analysis tools. With respect to MIAME, we concentrate on defining the content and structure of the necessary information rather than the technical format for capturing it.

Keywords: Computational Biology; Gene Expression Profiling, methods; Oligonucleotide Array Sequence Analysis, standards

Several programs (Catalyst, Confort, Flo99, MacroModel, and Omega) that are commonly used to generate conformational ensembles have been tested for their ability to reproduce bioactive conformations. The ligands from thirty-two different ligand-protein complexes determined by high-resolution (< 2.0 A) X-ray crystallography have been analyzed. The Low-Mode Conformational Search method (with AMBER* and the GB/SA hydration model), as implemented in MacroModel, was found to perform better than the other algorithms. The rule-based method Omega, which is orders of magnitude faster than the other methods, also gave reasonable results but were found to be dependent on the input structure. The methods supporting diverse sampling (Catalyst, Confort) performed least well. For the seven ligands in the set having eight or more rotatable bonds, none of the bioactive conformations were ever found, save for one exception (Flo99). These ligands do not bind in a local minimum conformation according to AMBER*/SA. Taking these last two observations together, it is clear that geometrically similar structures should be collected in order to increase the probability of finding the bioactive conformation among the generated ensembles. Factors influencing bioactive conformational retrieval have been identified and are discussed.

Keywords: Algorithms; Crystallography, X-Ray; Ligands; Models, Molecular; Molecular Conformation; Protein Binding; Quantum Theory; Software

Induced fit explains why biomolecules can bind together even if they are not optimized for binding. However, induced fit can lead to a kinetic bottleneck and does not describe every interaction in the absence of prior complementarity. Preselection of a fitting conformer is an alternative to induced fit.

Keywords: Antigen-Antibody Complex, physiology; Biological Products, chemistry/metabolism; Models, Biological; Molecular Conformation

Keywords: biosvm

We recently identified activating mutations of fibroblast growth factor receptor 3 (FGFR3) in bladder carcinoma. In this study we assessed the incidence of FGFR3 mutations in a series of 132 bladder carcinomas: 20 carcinoma in situ (CIS), 50 pTa, 19 pT1, and 43 pT2-4. All 48 mutations identified were identical to the germinal activating mutations that cause thanatophoric dysplasia, a lethal form of dwarfism. The S249C mutation, found in 33 of the 48 mutated tumors, was the most common. The frequency of mutations was higher in pTa tumors (37 of 50, 74 P < 0.0001) and pT2-4 tumors (7 of 43, 16 were detected in 27 of 32 (84 (7 grade was highly significant (P < 0.0001). FGFR3 is the first gene found to be mutated at a high frequency in pTa tumors. The absence of FGFR3 mutations in CIS and the low frequency of FGFR3 mutations in pT1 and pT2-4 tumors are consistent with the model of bladder tumor progression in which the most common precursor of pT1 and pT2-4 tumors is CIS.

We have generated a molecular taxonomy of lung carcinoma, the leading cause of cancer death in the United States and worldwide. Using oligonucleotide microarrays, we analyzed mRNA expression levels corresponding to 12,600 transcript sequences in 186 lung tumor samples, including 139 adenocarcinomas resected from the lung. Hierarchical and probabilistic clustering of expression data defined distinct subclasses of lung adenocarcinoma. Among these were tumors with high relative expression of neuroendocrine genes and of type II pneumocyte genes, respectively. Retrospective analysis revealed a less favorable outcome for the adenocarcinomas with neuroendocrine gene expression. The diagnostic potential of expression profiling is emphasized by its ability to discriminate primary lung adenocarcinomas from metastases of extra-pulmonary origin. These results suggest that integration of expression profile data with clinical parameters could aid in diagnosis of lung cancer patients.

Rapid accumulation of resistance mutations in the genome of the human immunodeficiency virus (HIV) plays a central role in drug treatment failure in infected patients. The authors have developed geno2pheno, an intelligent system that uses the information encoded in the viral genomic sequence to predict resistance or susceptibility of the virus to 13 antiretroviral agents. To predict phenotypic drug resistance from genotype, they applied two machine learning techniques: decision trees and linear support vector machines. These techniques performed learning on more than 400 genotype-phenotype pairs for each drug. The authors compared the generalization performance of the two families of models in leave-one-out experiments. Except for three drugs, all error estimates ranged between 7.25 and 15.5 percent. Support vector machines performed slightly better for most drugs, but knowledge extraction was easier for decision trees. Geno2pheno is freely available at http://cartan.gmd.de/geno2pheno.html.

Keywords: biosvm

In this paper we investigate the feasibility of using an SVM (support vector machine) classifier in our automatic system for the detection of clustered microcalcifications in digital mammograms. SVM is a technique for pattern recognition which relies on the statistical learning theory. It minimizes a function of two terms: the number of misclassified vectors of the training set and a term regarding the generalization classifier capability. We compare the SVM classifier with an MLP (multi-layer perceptron) in the false-positive reduction phase of our detection scheme: a detected signal is considered either microcalcification or false signal, according to the value of a set of its features. The SVM classifier gets slightly better results than the MLP one (Az value of 0.963 against 0.958) in the presence of a high number of training data; the improvement becomes much more evident (Az value of 0.952 against 0.918) in training sets of reduced size. Finally, the setting of the SVM classifier is much easier than the MLP one.

Keywords: biosvm image

G protein-coupled receptors (GPCRs) are a functionally diverse group of membrane proteins that play a critical role in signal transduction. Because of the lack of a high-resolution structure, the heptahelical transmembrane bundle within the N-terminal extracellular and C-terminal intracellular region of these receptors has initially been modeled based on the high-resolution structure of bacterial retinal-binding protein, bacteriorhodopsin. However, the low-resolution structure of rhodopsin, a prototypical GPCR, revealed that there is a minor relationship between GPCRs and bacteriorhodopsins. The high-resolution crystal structure of the rhodopsin ground state and further refinements of the model provide the first structural information about the entire organization of the polypeptide chain and post-translational moieties. These studies provide a structural template for Family 1 GPCRs that has the potential to significantly improve structure-based approaches to GPCR drug discovery.

Keywords: Amino Acid Sequence; Animals; Crystallography, X-Ray; Drug Design; GTP-Binding Proteins; Humans; Models, Molecular; Molecular Sequence Data; Receptors, Drug; Rhodopsin

Keywords: chemoinformatics

The essential genes, proteins and associated regulatory networks involved in the entry into the mammalian cell cycle are identified, from activation of growth-factor receptors to intracellular signal transduction pathways that impinge on the cell cycle machinery and ultimately on the initiation of DNA replication. Signaling pathways mediated by the oncoproteins Ras and Myc induce the activation of cyclin-dependent kinases CDK4 and CDK2, and the assembly and firing of pre-replication complexes require a collaboration among E2F, CDK2, and Cdc7 kinase. A proposed core mechanism of the restriction point, the major checkpoint prior to commitment to DNA synthesis, involves cyclin E/CDK2, the phosphatase Cdc25A, and the CDK inhibitor p27Kip1. (c) 2001 American Institute of Physics.

Motivation: The eXtensible Markup Language (XML) is an emerging standard for structuring documents, notably for the World Wide Web. In this paper, the authors present XML and examine its use as a data language for bioinformatics. In particular, XML is compared to other languages, and some of the potential uses of XML in bioinformatics applications are presented. The authors propose to adopt XML for data interchange between databases and other sources of data. Finally the discussion is illustrated by a test case of a pedigree data model in XML. Contact: Emmanuel.Barillot@infobiogen.fr

Keywords: Computational Biology; Humans; Information Storage and Retrieval; Internet; Programming Languages

Keywords: biosvm

Using methods of statistical physics, we investigate the role of model complexity in learning with support vector machines (SVMs), which are an important alternative to neural networks. We show the advantages of using SVMs with kernels of infinite complexity on noisy target rules, which, in contrast to common theoretical beliefs, are found to achieve optimal generalization error although the training error does not converge to the generalization error. Moreover, we find a universal asymptotics of the learning curves which depend only on the target rule but not on the SVM kernel.

Keywords: Algorithms, Amino Acid Sequence, Artificial Intelligence, Biological, Cell Compartmentation, Chemistry, Comparative Study, Computational Biology, Computer Simulation, Computer-Assisted, Databases, Decision Trees, Diagnosis, Discriminant Analysis, Electrophysiology, Factual, Gastric Emptying, Humans, Logistic Models, Melanoma, Models, Neural Networks (Computer), Nevus, Non-U.S. Gov't, Organelles, P.H.S., Physical, Pigmented, Predictive Value of Tests, Proteins, Proteome, Reproducibility of Results, Research Support, Skin Diseases, Skin Neoplasms, Skin Pigmentation, Software, Stomach Diseases, U.S. Gov't, 11328187

A recent study reported a conventional neural network (NN) approach for the noninvasive diagnosis of delayed gastric emptying from the cutaneous electrogastrograms. Using support vector machine, we show that this relatively new technique can be used for detection of delayed gastric emptying and is in fact able to outdo the conventional NN.

Keywords: Algorithms, Amino Acid Sequence, Artificial Intelligence, Biological, Cell Compartmentation, Comparative Study, Computer Simulation, Computer-Assisted, Decision Trees, Diagnosis, Discriminant Analysis, Electrophysiology, Gastric Emptying, Humans, Logistic Models, Melanoma, Models, Neural Networks (Computer), Nevus, Non-U.S. Gov't, Organelles, P.H.S., Pigmented, Predictive Value of Tests, Proteins, Reproducibility of Results, Research Support, Skin Diseases, Skin Neoplasms, Skin Pigmentation, Stomach Diseases, U.S. Gov't, 11341535

Keywords: Fungal Proteins, genetics/physiology; Gene Deletion; Protein Binding; Proteome; Saccharomyces cerevisiae, genetics/physiology; Signal Transduction

RNA interference (RNAi) is the process of sequence-specific, post-transcriptional gene silencing in animals and plants, initiated by double-stranded RNA (dsRNA) that is homologous in sequence to the silenced gene. The mediators of sequence-specific messenger RNA degradation are 21- and 22-nucleotide small interfering RNAs (siRNAs) generated by ribonuclease III cleavage from longer dsRNAs. Here we show that 21-nucleotide siRNA duplexes specifically suppress expression of endogenous and heterologous genes in different mammalian cell lines, including human embryonic kidney (293) and HeLa cells. Therefore, 21-nucleotide siRNA duplexes provide a new tool for studying gene function in mammalian cells and may eventually be used as gene-specific therapeutics.

Keywords: sirna

Functional annotation of newly sequenced genomes is an important challenge for computational biology systems. While much progress has been made towards scalingup experimental methods for functional assignment to putative genes, most current genomic annotation systems rely on computational solutions for homology modeling via sequence or structural similarity. We present a new method for remote homology detection that relies on combining probabilistic modeling and supervised learning in high-dimensional features spaces. Our system uses a transformation that converts protein domains to fixed-dimension representative feature vectors, where each feature records the sensitivity of each protein domain to a previously learned set of ?protein motifs? or ?blocks?. Subsequently, the system utilizes Support Vector Machine (SVM) classifiers to learn the boundaries between structural protein classes. Our experiments suggest that this technique performs well relative to several other remote homology methods for the majority of protein domains in SCOP 1.37 PDB90.

Keywords: biosvm

Keywords: biosvm chemoinformatics

The covering number of a ball of a reproducing kernel Hilbert space as a subset of the continuous function space plays an important role in Learning Theory. We give estimates for this covering number by means of the regularity of the Mercer kernel K. For convolution type kernels K(x, t) = k(x - t) on [0, 1]n, we provide estimates depending on the decay of k, the Fourier transform of k. In particular, when k decays exponentially, our estimate for this covering number is better than all the previous results and covers many important Mercer kernels. A counter example is presented to show that the eigenfunctions of the Hilbert-Schmidt operator LK associated with a Mercer kernel K may not be uniformly bounded. Hence some previous methods used for estimating the covering number in Learning Theory are not valid. We also provide an example of a Mercer kernel to show that LK½ may not be generated by a Mercer kernel.

Motivation: Data that characterize primary and tertiary structures of proteins are now accumulating at a rapid and accelerating rate and require automated computational tools to extract critical information relating amino acid changes with the spectrum of functionally attributes exhibited by a protein. We propose that immunoglobulin-type beta-domains, which are found in approximate 400 functionally distinct forms in humans alone, provide the immense genetic variation within limited conformational changes that might facilitate the development of new computational tools. As an initial step, we describe here an approach based on Support Vector Machine (SVM) technology to identify amino acid variations that contribute to the functional attribute of pathological self-assembly by some human antibody light chains produced during plasma cell diseases. Results: We demonstrate that SVMs with selective kernel scaling are an effective tool in discriminating between benign and pathologic human immunoglobulin light chains. Initial results compare favorably against manual classification performed by experts and indicate the capability of SVMs to capture the underlying structure of the data. The data set consists of 70 proteins of human antibody 1 light chains, each represented by aligned sequences of 120 amino acids. We perform feature selection based on a first-order adaptive scaling algorithm, which confirms the importance of changes in certain amino acid positions and identifies other positions that are key in the characterization of protein function.

Keywords: biosvm

A Support Vector Machine learning system has been trained to predict protein solvent accessibility from the primary structure. Different kernel functions and sliding window sizes have been explored to find how they affect the prediction performance. Using a cut-off threshold of 15 of exposed and buried residues), this method was able to achieve a prediction accuracy of 70.1 for multiple alignment sequence input, respectively. The prediction of three and more states of solvent accessibility was also studied and compared with other methods. The prediction accuracies are better than, or comparable to, those obtained by other methods such as neural networks, Bayesian classification, multiple linear regression, and information theory. In addition, our results further suggest that this system may be combined with other prediction methods to achieve more reliable results, and that the Support Vector Machine method is a very useful tool for biological sequence analysis.

Keywords: biosvm

BACKGROUND: A variety of methods for prediction of peptide binding to major histocompatibility complex (MHC) have been proposed. These methods are based on binding motifs, binding matrices, hidden Markov models (HMM), or artificial neural networks (ANN). There has been little prior work on the comparative analysis of these methods. MATERIALS AND METHODS: We performed a comparison of the performance of six methods applied to the prediction of two human MHC class I molecules, including binding matrices and motifs, ANNs, and HMMs. RESULTS: The selection of the optimal prediction method depends on the amount of available data (the number of peptides of known binding affinity to the MHC molecule of interest), the biases in the data set and the intended purpose of the prediction (screening of a single protein versus mass screening). When little or no peptide data are available, binding motifs are the most useful alternative to random guessing or use of a complete overlapping set of peptides for selection of candidate binders. As the number of known peptide binders increases, binding matrices and HMM become more useful predictors. ANN and HMM are the predictive methods of choice for MHC alleles with more than 100 known binding peptides. CONCLUSION: The ability of bioinformatic methods to reliably predict MHC binding peptides, and thereby potential T-cell epitopes, has major implications for clinical immunology, particularly in the area of vaccine design.

Keywords: Amino Acid Motifs; Computational Biology; Histocompatibility Antigens Class I; Humans; Models, Molecular; Peptides; Protein Binding

We propose a scheme to reverse-engineer gene networks on a genome-wide scale using a relatively small amount of gene expression data from microarray experiments. Our method is based on the empirical observation that such networks are typically large and sparse. It uses singular value decomposition to construct a family of candidate solutions and then uses robust regression to identify the solution with the smallest number of connections as the most likely solution. Our algorithm has O(log N) sampling complexity and O(N4) computational complexity. We test and validate our approach in a series of in numero experiments on model gene networks.

Keywords: Gene Expression Profiling; Gene Expression Regulation; Humans; Oligonucleotide Array Sequence Analysis, methods; Research Design; Time Factors

Keywords: chemoinformatics

The Database of Interacting Proteins (DIP: http://dip.doe-mbi.ucla.edu) is a database that documents experimentally determined protein-protein interactions. It provides the scientific community with an integrated set of tools for browsing and extracting information about protein interaction networks. As of September 2001, the DIP catalogs approximately 11 000 unique interactions among 5900 proteins from >80 organisms; the vast majority from yeast, Helicobacter pylori and human. Tools have been developed that allow users to analyze, visualize and integrate their own experimental data with the information about protein-protein interactions available in the DIP database.

Natural discriminant analysis based on interactive Potts models is developed in this work. A generative model composed of piece-wise multivariate gaussian distributions is used to characterize the input space, exploring the embedded clustering and mixing structures and developing proper internal representations of input parameters. The maximization of a log-likelihood function measuring the fitness of all input parameters to the generative model, and the minimization of a design cost summing up square errors between posterior outputs and desired outputs constitutes a mathematical framework for discriminant analysis. We apply a hybrid of the mean-field annealing and the gradient-descent methods to the optimization of this framework and obtain multiple sets of interactive dynamics, which realize coupled Potts models for discriminant analysis. The new learning process is a whole process of component analysis, clustering analysis, and labeling analysis. Its major improvement compared to the radial basis function and the support vector machine is described by using some artificial examples and a real-world application to breast cancer diagnosis.

Various unsupervised and supervised learning methods including support vector machines, classification trees, linear discriminant analysis and nearest neighbor classifiers have been used to classify high-throughput gene expression data. Simpler and more widely accepted statistical tools have not yet been used for this purpose, hence proper comparisons between classification methods have not been conducted. We developed free software that implements logistic regression with stepwise variable selection as a quick and simple method for initial exploration of important genetic markers in disease classification. To implement the algorithm and allow our collaborators in remote locations to evaluate and compare its results against those of other methods, we developed a user-friendly asynchronous web-based application with a minimal amount of programming using free, downloadable software tools. With this program, we show that classification using logistic regression can perform as well as other more sophisticated algorithms, and it has the advantages of being easy to interpret and reproduce. By making the tool freely and easily available, we hope to promote the comparison of classification methods. In addition, we believe our web application can be used as a model for other bioinformatics laboratories that need to develop web-based analysis tools in a short amount of time and on a limited budget.

Keywords: Acute, Algorithms, Animals, Artificial Intelligence, Automated, Base Pair Mismatch, Base Pairing, Base Sequence, Biological, Biosensing Techniques, Classification, Cluster Analysis, Comparative Study, Computational Biology, Computer-Assisted, Cystadenoma, DNA, Drug, Drug Design, Eukaryotic Cells, Female, Gene Expression, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Markers, Hemolysins, Humans, Internet, Leukemia, Ligands, Likelihood Functions, Logistic Models, Lymphocytic, Markov Chains, Mathematics, Messenger, Models, Molecular, Molecular Probe Techniques, Molecular Sequence Data, Nanotechnology, Neoplasm, Neoplasms, Neoplastic, Neural Networks (Computer), Non-P.H.S., Non-U.S. Gov't, Nucleic Acid Conformation, Observer Variation, Oligonucleotide Array Sequence Analysis, Ovarian Neoplasms, P.H.S., Pattern Recognition, Probability, Protein Binding, Proteins, Quality Control, RNA, RNA Splicing, Receptors, Reference Values, Reproducibility of Results, Research Support, Sensitivity and Specificity, Sequence Analysis, Signal Processing, Software, Statistical, Stomach Neoplasms, Thermodynamics, Transcription, Tumor Markers, U.S. Gov't, 12463949

The draft human genome sequence and the dissemination of high throughput technology provides opportunities for systematic analysis of cancer cells. Genome-wide mutation screens, high resolution analysis of chromosomal abberations and expression profiling all give comprehensive views of genetic alterations in cancer cells. From these analyses will come a complete list of the genetic changes that drive malignant transformation and of the therapeutic targets that may be exploited for clinical benefit.

Keywords: biosvm

Reproducing kernel Hilbert space (RKHS) methods provide a unified context for solving a wide variety of statistical modelling and function estimation problems. We consider two such problems: We are given a training set [yi, ti, i = 1, em leader, n], where yi is the response for the ith subject, and ti is a vector of attributes for this subject. The value of y(i) is a label that indicates which category it came from. For the first problem, we wish to build a model from the training set that assigns to each t in an attribute domain of interest an estimate of the probability pj(t) that a (future) subject with attribute vector t is in category j. The second problem is in some sense less ambitious; it is to build a model that assigns to each t a label, which classifies a future subject with that t into one of the categories or possibly "none of the above." The approach to the first of these two problems discussed here is a special case of what is known as penalized likelihood estimation. The approach to the second problem is known as the support vector machine. We also note some alternate but closely related approaches to the second problem. These approaches are all obtained as solutions to optimization problems in RKHS. Many other problems, in particular the solution of ill-posed inverse problems, can be obtained as solutions to optimization problems in RKHS and are mentioned in passing. We caution the reader that although a large literature exists in all of these topics, in this inaugural article we are selectively highlighting work of the author, former students, and other collaborators.

Keywords: Acute, Algorithms, Animals, Automated, Base Pair Mismatch, Base Pairing, Base Sequence, Biological, Biosensing Techniques, Classification, Cluster Analysis, Comparative Study, Computational Biology, Computer-Assisted, Cystadenoma, DNA, Drug, Drug Design, Eukaryotic Cells, Female, Gene Expression, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Markers, Hemolysins, Humans, Leukemia, Ligands, Likelihood Functions, Lymphocytic, Markov Chains, Mathematics, Messenger, Models, Molecular, Molecular Probe Techniques, Molecular Sequence Data, Nanotechnology, Neoplasm, Neoplastic, Neural Networks (Computer), Non-P.H.S., Non-U.S. Gov't, Nucleic Acid Conformation, Observer Variation, Oligonucleotide Array Sequence Analysis, Ovarian Neoplasms, P.H.S., Pattern Recognition, Probability, Protein Binding, Proteins, Quality Control, RNA, RNA Splicing, Receptors, Reference Values, Reproducibility of Results, Research Support, Sensitivity and Specificity, Sequence Analysis, Signal Processing, Statistical, Stomach Neoplasms, Thermodynamics, Transcription, Tumor Markers, U.S. Gov't, 12477931

BACKGROUND: A more accurate means of prognostication in breast cancer will improve the selection of patients for adjuvant systemic therapy. METHODS: Using microarray analysis to evaluate our previously established 70-gene prognosis profile, we classified a series of 295 consecutive patients with primary breast carcinomas as having a gene-expression signature associated with either a poor prognosis or a good prognosis. All patients had stage I or II breast cancer and were younger than 53 years old; 151 had lymph-node-negative disease, and 144 had lymph-node-positive disease. We evaluated the predictive power of the prognosis profile using univariable and multivariable statistical analyses. RESULTS: Among the 295 patients, 180 had a poor-prognosis signature and 115 had a good-prognosis signature, and the mean (+/-SE) overall 10-year survival rates were 54.6+/-4.4 percent and 94.5+/-2.6 percent, respectively. At 10 years, the probability of remaining free of distant metastases was 50.6+/-4.5 percent in the group with a poor-prognosis signature and 85.2+/-4.3 percent in the group with a good-prognosis signature. The estimated hazard ratio for distant metastases in the group with a poor-prognosis signature, as compared with the group with the good-prognosis signature, was 5.1 (95 percent confidence interval, 2.9 to 9.0; P<0.001). This ratio remained significant when the groups were analyzed according to lymph-node status. Multivariable Cox regression analysis showed that the prognosis profile was a strong independent factor in predicting disease outcome. CONCLUSIONS: The gene-expression profile we studied is a more powerful predictor of the outcome of disease in young patients with breast cancer than standard systems based on clinical and histologic criteria.

Keywords: breastcancer, csbcbook, csbcbook-ch3

Keywords: biosvm

Keywords: biosvm

Keywords: biosvm

Breast cancer patients with the same stage of disease can have markedly different treatment responses and overall outcome. The strongest predictors for metastases (for example, lymph node status and histological grade) fail to classify accurately breast tumours according to their clinical behaviour. Chemotherapy or hormonal therapy reduces the risk of distant metastases by approximately one-third; however, 70-80% of patients receiving this treatment would have survived without it. None of the signatures of breast cancer gene expression reported to date allow for patient-tailored therapy strategies. Here we used DNA microarray analysis on primary breast tumours of 117 young patients, and applied supervised classification to identify a gene expression signature strongly predictive of a short interval to distant metastases ('poor prognosis' signature) in patients without tumour cells in local lymph nodes at diagnosis (lymph node negative). In addition, we established a signature that identifies tumours of BRCA1 carriers. The poor prognosis signature consists of genes regulating cell cycle, invasion, metastasis and angiogenesis. This gene expression profile will outperform all currently used clinical parameters in predicting disease outcome. Our findings provide a strategy to select patients who would benefit from adjuvant therapy.

Keywords: breastcancer, csbcbook, csbcbook-ch3

Oral bioavailability measurements in rats for over 1100 drug candidates studied at SmithKline Beecham Pharmaceuticals (now GlaxoSmithKline) have allowed us to analyze the relative importance of molecular properties considered to influence that drug property. Reduced molecular flexibility, as measured by the number of rotatable bonds, and low polar surface area or total hydrogen bond count (sum of donors and acceptors) are found to be important predictors of good oral bioavailability, independent of molecular weight. That on average both the number of rotatable bonds and polar surface area or hydrogen bond count tend to increase with molecular weight may in part explain the success of the molecular weight parameter in predicting oral bioavailability. The commonly applied molecular weight cutoff at 500 does not itself significantly separate compounds with poor oral bioavailability from those with acceptable values in this extensive data set. Our observations suggest that compounds which meet only the two criteria of (1) 10 or fewer rotatable bonds and (2) polar surface area equal to or less than 140 A(2) (or 12 or fewer H-bond donors and acceptors) will have a high probability of good oral bioavailability in the rat. Data sets for the artificial membrane permeation rate and for clearance in the rat were also examined. Reduced polar surface area correlates better with increased permeation rate than does lipophilicity (C log P), and increased rotatable bond count has a negative effect on the permeation rate. A threshold permeation rate is a prerequisite of oral bioavailability. The rotatable bond count does not correlate with the data examined here for the in vivo clearance rate in the rat.

Keywords: chemogenomics

The large amount of data generated by DNA microarrays was originally analysed using unsupervised methods, such as clustering or self-organizing maps. Recently supervised methods such as decision trees, dot-product support vector machines (SVM) and multi-layer perceptrons (MLP) have been applied in order to classify normal and tumoural tissues. We propose methods based on non-linear SVM with polynomial and Gaussian kernels, and output coding (OC) ensembles of learning machines to separate normal from malignant tissues, to classify different types of lymphoma and to analyse the role of sets of coordinately expressed genes in carcinogenic processes of lymphoid tissues. Using gene expression data from "Lymphochip", a specialised DNA microarray developed at Stanford University School of Medicine, we show that SVM can correctly separate normal from tumoural tissues, and OC ensembles can be successfully used to classify different types of lymphoma. Moreover, we identify a group of coordinately expressed genes related to the separation of two distinct subgroups inside diffuse large B-cell lymphoma (DLBCL), validating a previous Alizadeh's hypothesis about the existence of two distinct diseases inside DLBCL.

Keywords: biosvm

Keywords: csbcbook

Escherichia coli MG1655 acid-inducible genes were identified by whole-genome expression profiling. Cultures were grown to the mid-logarithmic phase on acidified glucose minimal medium, conditions that induce glutamate-dependent acid resistance (AR), while the other AR systems are either repressed or not induced. A total of 28 genes were induced in at least two of three experiments in which the gene expression profiles of cells grown in acid (pH 5.5 or 4.5) were compared to those of cells grown at pH 7.4. As expected, the genes encoding glutamate decarboxylase, gadA and gadB, were significantly induced. Interestingly, two acid-inducible genes code for small basic proteins with pIs of >10.5, and six code for small acidic proteins with pIs ranging from 5.7 to 4.0; the roles of these small basic and acidic proteins in acid resistance are unknown. The acid-induced genes represented only five functional grouping categories, including eight genes involved in metabolism, nine associated with cell envelope structures or modifications, two encoding chaperones, six regulatory genes, and six unknown genes. It is unlikely that all of these genes are involved in the glutamate-dependent AR. However, nine acid-inducible genes are clustered in the gadA region, including hdeA, which encodes a putative periplasmic chaperone, and four putative regulatory genes. One of these putative regulators, yhiE, was shown to significantly increase acid resistance when overexpressed in cells that had not been preinduced by growth at pH 5.5, and mutation of yhiE decreased acid resistance; yhiE could therefore encode an activator of AR genes. Thus, the acid-inducible genes clustered in the gadA region appear to be involved in glutatmate-dependent acid resistance, although their specific roles remain to be elucidated.

Keywords: Culture Media; Escherichia coli; Escherichia coli Proteins; Gene Expression Profiling; Gene Expression Regulation, Bacterial; Heat-Shock Response; Hydrogen-Ion Concentration; Morpholines; Oligonucleotide Array Sequence Analysis

Motivation: Kernel methods such as support vector machines require a kernel function between objects to be defined a priori. Several works have been done to derive kernels from probability distributions, e.g., the Fisher kernel. However, a general methodology to design a kernel is not fully developed. Results: We propose a reasonable way of designing a kernel when objects are generated from latent variable models (e.g., HMM). First of all, a joint kernel is designed for complete data which include both visible and hidden variables. Then a marginalized kernel for visible data is obtained by taking the expectation with respect to hidden variables. We will show that the Fisher kernel is a special case of marginalized kernels, which gives another viewpoint to the Fisher kernel theory. Although our approach can be applied to any object, we particularly derive several marginalized kernels useful for biological sequences (e.g., DNA and proteins). The effectiveness of marginalized kernels is illustrated in the task of classifying bacterial gyrase subunit B (gyrB) amino acid sequences.

Keywords: biosvm

Keywords: biosvm

Keywords: chemoinformatics

Human immunodeficiency virus type 1 (HIV-1) gene expression is regulated by both cellular transcription factors and Tat. The ability of Tat to stimulate transcriptional elongation is dependent on its binding to TAR RNA in conjunction with cyclin T1 and CDK9. A variety of other cellular factors that bind to the HIV-1 long terminal repeat, including NF-kappaB, SP1, LBP, and LEF, are also important in the control of HIV-1 gene expression. Although these factors have been demonstrated to regulate HIV-1 gene expression by both genetic and biochemical analysis, in most cases a direct in vivo demonstration of their role on HIV-1 replication has not been established. Recently, the efficacy of RNA interference in mammalian cells has been shown utilizing small interfering RNAs (siRNAs) to result in the specific degradation of host mRNAs and decreases the levels of their corresponding proteins. In this study, we addressed whether siRNAs directed against either HIV-1 tat or reverse transcriptase or the NF-kappaB p65 subunit could specifically decrease the levels of these proteins and thus alter HIV-1 replication. Our results demonstrate the specificity of siRNAs for decreasing the expression of these viral and cellular proteins and inhibiting HIV-1 replication. These studies suggest that RNA interference is useful in exploring the biological role of cellular and viral regulatory factors involved in the control of HIV-1 gene expression.

Keywords: sirna

MOTIVATION: With the increasing number of gene expression databases, the need for more powerful analysis and visualization tools is growing. Many techniques have successfully been applied to unravel latent similarities among genes and/or experiments. Most of the current systems for microarray data analysis use statistical methods, hierarchical clustering, self-organizing maps, support vector machines, or k-means clustering to organize genes or experiments into 'meaningful' groups. Without prior explicit bias almost all of these clustering methods applied to gene expression data not only produce different results, but may also produce clusters with little or no biological relevance. Of these methods, agglomerative hierarchical clustering has been the most widely applied, although many limitations have been identified. RESULTS: Starting with a systematic comparison of the underlying theories behind clustering approaches, we have devised a technique that combines tree-structured vector quantization and partitive k-means clustering (BTSVQ). This hybrid technique has revealed clinically relevant clusters in three large publicly available data sets. In contrast to existing systems, our approach is less sensitive to data preprocessing and data normalization. In addition, the clustering results produced by the technique have strong similarities to those of self-organizing maps (SOMs). We discuss the advantages and the mathematical reasoning behind our approach.

High-throughput gene expression profiling has become an important tool for investigating transcriptional activity in a variety of biological samples. To date, the vast majority of these experiments have focused on specific biological processes and perturbations. Here, we have generated and analyzed gene expression from a set of samples spanning a broad range of biological conditions. Specifically, we profiled gene expression from 91 human and mouse samples across a diverse array of tissues, organs, and cell lines. Because these samples predominantly come from the normal physiological state in the human and mouse, this dataset represents a preliminary, but substantial, description of the normal mammalian transcriptome. We have used this dataset to illustrate methods of mining these data, and to reveal insights into molecular and physiological gene function, mechanisms of transcriptional regulation, disease etiology, and comparative genomics. Finally, to allow the scientific community to use this resource, we have built a free and publicly accessible website (http://expression.gnf.org) that integrates data visualization and curation of current gene annotations.

Keywords: Animals; Collagen; Female; Gene Expression Profiling; Humans; Male; Mice; Organ Specificity; Polymerase Chain Reaction; Receptors, Cell Surface; Transcription, Genetic

A versatile, platform independent and easy to use Java suite for large-scale gene expression analysis was developed. Genesis integrates various tools for microarray data analysis such as filters, normalization and visualization tools, distance measures as well as common clustering algorithms including hierarchical clustering, self-organizing maps, k-means, principal component analysis, and support vector machines. The results of the clustering are transparent across all implemented methods and enable the analysis of the outcome of different algorithms and parameters. Additionally, mapping of gene expression data onto chromosomal sequences was implemented to enhance promoter analysis and investigation of transcriptional control mechanisms.

Keywords: Algorithms, Artificial Intelligence, Cluster Analysis, Comparative Study, Computational Biology, Databases, Gene Expression Profiling, Genetic, Models, Molecular Structure, Neural Networks (Computer), Non-U.S. Gov't, Oligonucleotide Array Sequence Analysis, Principal Component Analysis, Programming Languages, Promoter Regions (Genetics), Protein, Proteins, Research Support, Software, Statistical, Transcription, 11836235

Recent studies of histone methylation have yielded fundamental new insights pertaining to the role of this modification in gene activation as well as in gene silencing. While a number of methylation sites are known to occur on histones, only limited information exists regarding the relevant enzymes that mediate these methylation events. We thus sought to identify native histone methyltransferase (HMT) activities from Saccharomyces cerevisiae. Here, we describe the biochemical purification and characterization of Set2, a novel HMT that is site-specific for lysine 36 (Lys36) of the H3 tail. Using an antiserum directed against Lys36 methylation in H3, we show that Set2, via its SET domain, is responsible for methylation at this site in vivo. Tethering of Set2 to a heterologous promoter reveals that Set2 represses transcription, and part of this repression is mediated through the HMT activity of the SET domain. These results suggest that Set2 and methylation at H3 Lys36 play a role in the repression of gene transcription.

Keywords: Amino Acid Sequence; Gene Expression Regulation, Fungal; Histones, metabolism; Methyltransferases, metabolism; Molecular Sequence Data; Nucleosomes, enzymology; Saccharomyces cerevisiae Proteins, metabolism; Saccharomyces cerevisiae, enzymology/genetics; Substrate Specificity; Transcription, Genetic; Transcriptional Activation

The relationship between structure, function and regulation in complex cellular networks is a still largely open question. Systems biology aims to explain this relationship by combining experimental and theoretical approaches. Current theories have various strengths and shortcomings in providing an integrated, predictive description of cellular networks. Specifically, dynamic mathematical modelling of large-scale networks meets difficulties because the necessary mechanistic detail and kinetic parameters are rarely available. In contrast, structure-oriented analyses only require network topology, which is well known in many cases. Previous approaches of this type focus on network robustness or metabolic phenotype, but do not give predictions on cellular regulation. Here, we devise a theoretical method for simultaneously predicting key aspects of network functionality, robustness and gene regulation from network structure alone. This is achieved by determining and analysing the non-decomposable pathways able to operate coherently at steady state (elementary flux modes). We use the example of Escherichia coli central metabolism to illustrate the method.

We show that support vector machines of the 1-norm soft margin type are universally consistent provided that the regularization parameter is chosen in a distinct manner and the kernel belongs to a specific class?the so-called universal kernels?which has recently been considered by the author. In particular it is shown that the 1-norm soft margin classifier with Gaussian RBF kernel on a compact subset X of d and regularization parameter cn=n??1 is universally consistent, if n is the training set size and 0<?<1/d.

We present an automatic method to classify the sub-cellular location of proteins based on the text of relevant medline abstracts. For each protein, a vector of terms is generated from medline abstracts in which the protein/gene's name or synonym occurs. A Support Vector Machine (SVM) is used to automatically partition the term space and to thus discriminate the textual features that define sub-cellular location. The method is benchmarked on a set of proteins of known sub-cellular location from S. cerevisiae. No prior knowledge of the problem domain nor any natural language processing is used at any stage. The method out-performs support vector machines trained on amino acid composition and has comparable performance to rule-based text classifiers. Combining text with protein amino-acid composition improves recall for some sub-cellular locations. We discuss the generality of the method and its potential application to a variety of biological classification problems.

Keywords: biosvm

Keywords: csbcbook

Keywords: biosvm

Quantitative Structure-Retention Relationship (QSRR) models are developed for the prediction of protein retention times in anion-exchange chromatography systems. Topological, subdivided surface area, and TAE (Transferable Atom Equivalent) electron-density-based descriptors are computed directly for a set of proteins using molecular connectivity patterns and crystal structure geometries. A novel algorithm based on Support Vector Machine (SVM) regression has been employed to obtain predictive QSRR models using a two-step computational strategy. In the first step, a sparse linear SVM was utilized as a feature selection procedure to remove irrelevant or redundant information. Subsequently, the selected features were used to produce an ensemble of nonlinear SVM regression models that were combined using bootstrap aggregation (bagging) techniques, where various combinations of training and validation data sets were selected from the pool of available data. A visualization scheme (star plots) was used to display the relative importance of each selected descriptor in the final set of "bagged" models. Once these predictive models have been validated, they can be used as an automated prediction tool for virtual high-throughput screening (VHTS).

Keywords: Acute, Algorithms, Animals, Anion Exchange Resins, Artificial Intelligence, Automated, Base Pair Mismatch, Base Pairing, Base Sequence, Biological, Biosensing Techniques, Carcinoma, Chemical, Chromatography, Classification, Cluster Analysis, Comparative Study, Computational Biology, Computer-Assisted, Cystadenoma, DNA, Decision Making, Diagnosis, Differential, Drug, Drug Design, Electrostatics, Eukaryotic Cells, Feasibility Studies, Female, Gene Expression, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Markers, Hemolysins, Humans, Internet, Ion Exchange, Leukemia, Ligands, Likelihood Functions, Logistic Models, Lung Neoplasms, Lymphocytic, Lymphoma, Markov Chains, Mathematics, Messenger, Models, Molecular, Molecular Probe Techniques, Molecular Sequence Data, Nanotechnology, Neoplasm, Neoplasms, Neoplastic, Neural Networks (Computer), Non-P.H.S., Non-Small-Cell Lung, Non-U.S. Gov't, Nucleic Acid Conformation, Nucleic Acid Hybridization, Observer Variation, Oligonucleotide Array Sequence Analysis, Ovarian Neoplasms, P.H.S., Pattern Recognition, Probability, Protein Binding, Protein Conformation, Proteins, Quality Control, Quantum Theory, RNA, RNA Splicing, Receptors, Reference Values, Regression Analysis, Reproducibility of Results, Research Support, Sensitivity and Specificity, Sequence Analysis, Signal Processing, Software, Statistical, Stomach Neoplasms, Thermodynamics, Transcription, Tumor Markers, U.S. Gov't, 12444731

Many different biological questions are routinely studied using transcriptional profiling on microarrays. A wide range of approaches are available for gleaning insights from the data obtained from such experiments. The appropriate choice of data-analysis technique depends both on the data and on the goals of the experiment. This review summarizes some of the common themes in microarray data analysis, including detection of differential expression, clustering, and predicting sample characteristics. Several approaches to each problem, and their relative merits, are discussed and key areas for additional research highlighted.

Diffuse large B-cell lymphoma (DLBCL), the most common lymphoid malignancy in adults, is curable in less than 50 models based on pre-treatment characteristics, such as the International Prognostic Index (IPI), are currently used to predict outcome in DLBCL. However, clinical outcome models identify neither the molecular basis of clinical heterogeneity, nor specific therapeutic targets. We analyzed the expression of 6,817 genes in diagnostic tumor specimens from DLBCL patients who received cyclophosphamide, adriamycin, vincristine and prednisone (CHOP)-based chemotherapy, and applied a supervised learning prediction method to identify cured versus fatal or refractory disease. The algorithm classified two categories of patients with very different five-year overall survival rates (70 within specific IPI risk categories who were likely to be cured or to die of their disease. Genes implicated in DLBCL outcome included some that regulate responses to B-cell?receptor signaling, critical serine/threonine phosphorylation pathways and apoptosis. Our data indicate that supervised learning classification techniques can predict outcome in DLBCL and identify rational targets for intervention.

Keywords: biosvm

An important goal of whole-cell computational modeling is to integrate detailed biochemical information with biological intuition to produce testable predictions. Based on the premise that prokaryotes such as Escherichia coli have maximized their growth performance along evolution, flux balance analysis (FBA) predicts metabolic flux distributions at steady state by using linear programming. Corroborating earlier results, we show that recent intracellular flux data for wild-type E. coli JM101 display excellent agreement with FBA predictions. Although the assumption of optimality for a wild-type bacterium is justifiable, the same argument may not be valid for genetically engineered knockouts or other bacterial strains that were not exposed to long-term evolutionary pressure. We address this point by introducing the method of minimization of metabolic adjustment (MOMA), whereby we test the hypothesis that knockout metabolic fluxes undergo a minimal redistribution with respect to the flux configuration of the wild type. MOMA employs quadratic programming to identify a point in flux space, which is closest to the wild-type point, compatibly with the gene deletion constraint. Comparing MOMA and FBA predictions to experimental flux data for E. coli pyruvate kinase mutant PB25, we find that MOMA displays a significantly higher correlation than FBA. Our method is further supported by experimental data for E. coli knockout growth rates. It can therefore be used for predicting the behavior of perturbed metabolic networks, whose growth performance is in general suboptimal. MOMA and its possible future extensions may be useful in understanding the evolutionary optimization of metabolism.

Keywords: biosvm

Elementary flux modes (direct reaction routes) are minimal sets of enzymes that can operate at steady state, with all irreversible reactions used in the appropriate direction. They can be interpreted as component pathways of a (bio)chemical reaction network. Here, two different definitions of elementary modes are given and their equivalence is proved. Several algebraic properties of elementary modes are then presented and proved. This concerns, amongst other features, the minimal number of enzymes of the network not used in an elementary mode and the situations where irreversible reactions are replaced by reversible ones. Based on these properties, a refined algorithm is presented, and it is formally proved that this algorithm will exclusively generate all the elementary flux modes of an arbitrary network containing reversible or irreversible reactions or both. The algorithm is illustrated by a biochemical example relevant in nucleotide metabolism. The computer implementation in two different programming languages is discussed.

Annotation efforts in biosciences have focused in past years mainly on the annotation of genomic sequences. Only very limited effort has been put into annotation schemes for pharmaceutical ligands. Here we propose annotation schemes for the ligands of four major target classes, enzymes, G protein-coupled receptors (GPCRs), nuclear receptors (NRs), and ligand-gated ion channels (LGICs), and outline their usage for in silico screening and combinatorial library design. The proposed schemes cover ligand functionality and hierarchical levels of target classification. The classification schemes are based on those established by the EC, GPCRDB, NuclearDB, and LGICDB. The ligands of the MDL Drug Data Report (MDDR) database serve as a reference data set of known pharmacologically active compounds. All ligands were annotated according to the schemes when attribution was possible based on the activity classification provided by the reference database. The purpose of the ligand-target classification schemes is to allow annotation-based searching of the ligand database. In addition, the biological sequence information of the target is directly linkable to the ligand, hereby allowing sequence similarity-based identification of ligands of next homologous receptors. Ligands of specified levels can easily be retrieved to serve as comprehensive reference sets for cheminformatics-based similarity searches and for design of target class focused compound libraries. Retrospective in silico screening experiments within the MDDR01.1 database, searching for structures binding to dopamine D2, all dopamine receptors and all amine-binding class A GPCRs using known dopamine D2 binding compounds as a reference set, have shown that such reference sets are in particular useful for the identification of ligands binding to receptors closely related to the reference system. The potential for ligand identification drops with increasing phylogenetic distance. The analysis of the focus of a tertiary amine based combinatorial library compared to known amine binding class A GPCRs, peptide binding class A GPCRs, and LGIC ligands constitutes a second application scenario which illustrates how the focus of a combinatorial library can be treated quantitatively. The provided annotation schemes, which bridge chem- and bioinformatics by linking ligands to sequences, are expected to be of key utility for further systematic chemogenomics exploration of previously well explored target families.

Keywords: chemogenomics

A new method has been developed to detect functional relationships among proteins independent of a given sequence or fold homology. It is based on the idea that protein function is intimately related to the recognition and subsequent response to the binding of a substrate or an endogenous ligand in a well-characterized binding pocket. Thus, recognition of similar ligands, supposedly linked to similar function, requires conserved recognition features exposed in terms of common physicochemical interaction properties via the functional groups of the residues flanking a particular binding cavity. Following a technique commonly used in the comparison of small molecule ligands, generic pseudocenters coding for possible interaction properties were assigned for a large sample set of cavities extracted from the entire PDB and stored in the database Cavbase. Using a particular query cavity a series of related cavities of decreasing similarity is detected based on a clique detection algorithm. The detected similarity is ranked according to property-based surface patches shared in common by the different clique solutions. The approach either retrieves protein cavities accommodating the same (e.g. co-factors) or closely related ligands or it extracts proteins exhibiting similar function in terms of a related catalytic mechanism. Finally the new method has strong potential to suggest alternative molecular skeletons in de novo design. The retrieval of molecular building blocks accommodated in a particular sub-pocket that shares similarity with the pocket in a protein studied by drug design can inspire the discovery of novel ligands.

Keywords: Algorithms; Binding Sites; Databases, Protein; Models, Molecular; Molecular Structure; Protein Binding; Protein Folding; Protein Structure, Tertiary; Proteins, chemistry/metabolism; Reproducibility of Results

PURPOSE: To compare the ability of several machine learning classifiers to predict development of abnormal fields at follow-up in ocular hypertensive (OHT) eyes that had normal visual fields in baseline examination. METHODS: The visual fields of 114 eyes of 114 patients with OHT with four or more visual field tests with standard automated perimetry over three or more years and for whom stereophotographs were available were assessed. The mean (+/-SD) number of visual field tests was 7.89 +/- 3.04. The mean number of years covered (+/-SD) was 5.92 +/- 2.34 (range, 2.81-11.77). Fields were classified as normal or abnormal based on Statpac-like methods (Humphrey Instruments, Dublin, CA) and by several machine learning classifiers. The machine learning classifiers were two types of support vector machine (SVM), a mixture of Gaussian (MoG) classifier, a constrained MoG, and a mixture of generalized Gaussian (MGG). Specificity was set to 96% for all classifiers, using data from 94 normal eyes evaluated longitudinally. Specificity cutoffs required confirmation of abnormality. RESULTS: Thirty-two percent (36/114) of the eyes converted to abnormal fields during follow-up based on the Statpac-like methods. All 36 were identified by at least one machine classifier. In nearly all cases, the machine learning classifiers predicted the confirmed abnormality, on average, 3.92 +/- 0.55 years earlier than traditional Statpac-like methods. CONCLUSIONS: Machine learning classifiers can learn complex patterns and trends in data and adapt to create a decision surface without the constraints imposed by statistical classifiers. This adaptation allowed the machine learning classifiers to identify abnormality in visual field converts much earlier than the traditional methods.

We predict regulatory targets for 14 Arabidopsis microRNAs (miRNAs) by identifying mRNAs with near complementarity. Complementary sites within predicted targets are conserved in rice. Of the 49 predicted targets, 34 are members of transcription factor gene families involved in developmental patterning or cell differentiation. The near-perfect complementarity between plant miRNAs and their targets suggests that many plant miRNAs act similarly to small interfering RNAs and direct mRNA cleavage. The targeting of developmental transcription factors suggests that many plant miRNAs function during cellular differentiation to clear key regulatory transcripts from daughter cell lineages.

Keywords: sirna

Peptides that bind to a given major histocompatibility complex (MHC) molecule share sequence similarity. Therefore, a position specific scoring matrix (PSSM) or profile derived from a set of peptides known to bind to a specific MHC molecule would be a suitable predictor of whether other peptides might bind, thus anticipating possible T-cell epitopes within a protein. In this approach, the binding potential of any peptide sequence (query) to a given MHC molecule is linked to its similarity to a group of aligned peptides known to bind to that MHC, and can be obtained by comparing the query to the PSSM. This article describes the derivation of alignments and profiles from a collection of peptides known to bind a specific MHC, compatible with the structural and molecular basis of the peptide-MHC class I (MHCI) interaction. Moreover, in order to apply these profiles to the prediction of peptide-MHCI binding, we have developed a new search algorithm (RANKPEP) that ranks all possible peptides from an input protein using the PSSM coefficients. The predictive power of the method was evaluated by running RANKPEP on proteins known to bear MHCI K(b)- and D(b)-restricted T-cell epitopes. Analysis of the results indicates that > 80% of these epitopes are among the top 2% of scoring peptides. Prediction of peptide-MHC binding using a variety of MHCI-specific PSSMs is available on line at our RANKPEP web server (www.mifoundation.org/Tools/rankpep.html). In addition, the RANKPEP server also allows the user to enter additional profiles, making the server a powerful and versatile computational biology benchmark for the prediction of peptide-MHC binding.

Keywords: immunoinformatics

Keywords: Animals; Data Interpretation, Statistical; Forecasting; Gene Expression Profiling, methods; Humans; Oligonucleotide Array Sequence Analysis, methods; Research Design

This pilot study was carried out to find the feasibility of analyzing the maturity of the fetal lung using ultrasound images. Data were collected from normal pregnant women at intervals of two weeks from the gestation age of 24 to 38 weeks. Images were acquired at two centers located at different geographical locations. The total data acquired consisted of 750 images of immature and 250 images of mature class. A region of interest of 64 x 64 pixels was used for extracting the features. Various textural features were computed from the fetal lung and liver images. The ratios of fetal lung to liver feature values were investigated as possible indexes for classifying the images into those from mature (reduced pulmonary risk) and immature (possible pulmonary risk) lung. The features used are fractal dimension, lacunarity, and features derived from the histogram of the images. The following classifiers were used to classify the fetal lung images as belonging to mature or immature lung: nearest neighbor, k-nearest neighbor, modified k-nearest neighbor, multilayer perceptron, radial basis function network, and support vector machines. The classification accuracy obtained for the testing set ranges from 73% to 96%.

Genomic DNA copy number alterations are key genetic events in the development and progression of human cancers. Here we report a genome-wide microarray comparative genomic hybridization (array CGH) analysis of DNA copy number variation in a series of primary human breast tumors. We have profiled DNA copy number alteration across 6,691 mapped human genes, in 44 predominantly advanced, primary breast tumors and 10 breast cancer cell lines. While the overall patterns of DNA amplification and deletion corroborate previous cytogenetic studies, the high-resolution (gene-by-gene) mapping of amplicon boundaries and the quantitative analysis of amplicon shape provide significant improvement in the localization of candidate oncogenes. Parallel microarray measurements of mRNA levels reveal the remarkable degree to which variation in gene copy number contributes to variation in gene expression in tumor cells. Specifically, we find that 62% of highly amplified genes show moderately or highly elevated expression, that DNA copy number influences gene expression across a wide range of DNA copy number alterations (deletion, low-, mid- and high-level amplification), that on average, a 2-fold change in DNA copy number is associated with a corresponding 1.5-fold change in mRNA levels, and that overall, at least 12% of all the variation in gene expression among the breast tumors is directly attributable to underlying variation in gene copy number. These findings provide evidence that widespread DNA copy number alteration can lead directly to global deregulation of gene expression, which may contribute to the development or progression of cancer.

Keywords: Breast Neoplasms, genetics; Chromosome Aberrations; Disease Progression; Gene Dosage; Genome; Humans; Oligonucleotide Array Sequence Analysis; RNA, Messenger, metabolism; Transcription, Genetic; Tumor Cells, Cultured

Although approximately one-quarter of the roughly 4,000 genetically inherited diseases currently recorded in respective databases (LocusLink, OMIM) are already linked to a region of the human genome, about 450 have no known associated gene. Finding disease-related genes requires laborious examination of hundreds of possible candidate genes (sometimes, these are not even annotated; see, for example, refs 3,4). The public availability of the human genome draft sequence has fostered new strategies to map molecular functional features of gene products to complex phenotypic descriptions, such as those of genetically inherited diseases. Owing to recent progress in the systematic annotation of genes using controlled vocabularies, we have developed a scoring system for the possible functional relationships of human genes to 455 genetically inherited diseases that have been mapped to chromosomal regions without assignment of a particular gene. In a benchmark of the system with 100 known disease-associated genes, the disease-associated gene was among the 8 best-scoring genes with a 25% chance, and among the best 30 genes with a 50% chance, showing that there is a relationship between the score of a gene and its likelihood of being associated with a particular disease. The scoring also indicates that for some diseases, the chance of identifying the underlying gene is higher.

In our attempts to understand cellular function at the molecular level, we must be able to synthesize information from disparate types of genomic data. We consider the problem of inferring gene functional classifications from a heterogeneous data set consisting of DNA microarray expression measurements and phylogenetic profiles from whole-genome sequence comparisons. We demonstrate the application of the support vector machine (SVM) learning algorithm to this functional inference task. Our results suggest the importance of exploiting prior information about the heterogeneity of the data. In particular, we propose an SVM kernel function that is explicitly heterogeneous. In addition, we describe feature scaling methods for further exploiting prior knowledge of heterogeneity by giving each data type different weights.

Keywords: biosvm

We address a commonly asked question about gene expression data sets: "What functional classes of genes are most interesting in the data?" In the methods we present, expression data is partitioned into classes based on existing annotation schemes. Each class is then given three separately derived "interest" scores. The first score is based on an assessment of the statistical significance of gene expression changes experienced by members of the class, in the context of the experimental design. The second is based on the co-expression of genes in the class. The third score is based on the learnability of the classification. We show that all three methods reveal significant classes in each of three different gene expression data sets. Many classes are identified by one method but not the others, indicating that the methods are complementary. The classes identified are in many cases of clear relevance to the experiment. Our results suggest that these class scoring methods are useful tools for exploring gene expression data.

Accurate splice site prediction is a critical component of any computational approach to gene prediction in higher organisms. Existing approaches generally use sequence-based models that capture local dependencies among nucleotides in a small window around the splice site. We present evidence that computationally predicted secondary structure of moderate length pre-mRNA subsequencies contains information that can be exploited to improve acceptor splice site prediction beyond that possible with conventional sequence-based approaches. Both decision tree and support vector machine classifiers, using folding energy and structure metrics characterizing helix formation near the splice site, achieve a 5-10 a human data set. Based on our data, we hypothesize that acceptors preferentially exhibit short helices at the splice site.

Keywords: biosvm

Motivation: In this paper we address the problem of the determination of developmental age of an embryo from its segmentation gene expression patterns in Drosophila. Results: By applying support vector regression we have developed a fast method for automated staging of an embryo on the basis of its gene expression pattern. Support vector regression is a statistical method for creating regression functions of arbitrary type from a set of training data. The training set is composed of embryos for which the precise developmental age was determined by measuring the degree of membrane invagination. Testing the quality of regression on the training set showed good prediction accuracy. The optimal regression function was then used for the prediction of the gene expression based age of embryos in which the precise age has not been measured by membrane morphology. Moreover, we show that the same accuracy of prediction can be achieved when the dimensionality of the feature vector was reduced by applying factor analysis. The data reduction allowed us to avoid over-fitting and to increase the efficiency of the algorithm. Availability: This software may be obtained from the authors. Contact: samson@fn.csa.ru Keywords: gene expression patterns; development; embryo staging; support vector regression; segmentation genes; Drosophila.

Keywords: biosvm

The very high dimensional space of gene expression measurements obtained by DNA microarrays impedes the detection of underlying patterns in gene expression data and the identification of discriminatory genes. In this paper we show the use of projection methods such as principal components analysis (PCA) to obtain a direct link between patterns in the genes and patterns in samples. This feature is useful in the initial interactive pattern exploration of gene expression data and data-driven learning of the nature and types of samples. Using oligonucleotide microarray measurements of 40 samples from different normal human tissues, we show that distinct patterns are obtained when the genes are projected on a two-dimensional plane spanned by the loadings of the two major principal components. These patterns define the particular genes associated with a sample class (i.e., tissue). When used separately from the other genes, these class-specific (i.e., tissue-specific) genes in turn define distinct tissue patterns in the projection space spanned by the scores of the two major principal components. In this study, PCA projection facilitated discriminatory gene selection for different tissues and identified tissue-specific gene expression signatures for liver, skeletal muscle, and brain samples. Furthermore, it allowed the classification of nine new samples belonging to these three types using the linear combination of the expression levels of the tissue-specific genes determined from the first set of samples. The application of the technique to other published data sets is also discussed.

The rationale for developing anti-HIV vaccines that stimulate cytotoxic T-lymphocyte responses is given. We argue that such vaccines will work, provided that attention is paid to the development of memory T-cell responses that are strong and preferably activated. Furthermore, the vaccine should match the prevailing virus clade as closely as possible. Vaccines will have to stimulate a wide range of responses, but it is not clear how this can be achieved.

Keywords: immunoinformatics

Among the 3 billion base pairs of the human genome, there are approximately 30,000-40,000 protein-coding genes, but the function of at least half of them remains unknown. A new tool - short interfering RNAs (siRNAs) - has now been developed for systematically deciphering the functions and interactions of these thousands of genes. siRNAs are an intermediate of RNA interference, the process by which double-stranded RNA silences homologous genes. Although the use of siRNAs to silence genes in vertebrate cells was only reported a year ago, the emerging literature indicates that most vertebrate genes can be studied with this technology.

Keywords: sirna

Recent advances in microarray technology have opened new ways for functional annotation of previously uncharacterised genes on a genomic scale. This has been demonstrated by unsupervised clustering of co-expressed genes and, more importantly, by supervised learning algorithms. Using prior knowledge, these algorithms can assign functional annotations based on more complex expression signatures found in existing functional classes. Previously, support vector machines (SVMs) and other machine-learning methods have been applied to a limited number of functional classes for this purpose. Here we present, for the first time, the comprehensive application of supervised neural networks (SNNs) for functional annotation. Our study is novel in that we report systematic results for  100 classes in the Munich Information Center for Protein Sequences (MIPS) functional catalog. We found that only  10% of these are learnable (based on the rate of false negatives). A closer analysis reveals that false positives (and negatives) in a machine-learning context are not necessarily "false" in a biological sense. We show that the high degree of interconnections among functional classes confounds the signatures that ought to be learned for a unique class. We term this the "Borges effect" and introduce two new numerical indices for its quantification. Our analysis indicates that classification systems with a lower Borges effect are better suitable for machine learning. Furthermore, we introduce a learning procedure for combining false positives with the original class. We show that in a few iterations this process converges to a gene set that is learnable with considerably low rates of false positives and negatives and contains genes that are biologically related to the original class, allowing for a coarse reconstruction of the interactions between associated biological pathways. We exemplify this methodology using the well-studied tricarboxylic acid cycle.

Keywords: Acute, Algorithms, Animals, Anion Exchange Resins, Artificial Intelligence, Automated, Base Pair Mismatch, Base Pairing, Base Sequence, Biological, Biosensing Techniques, Carcinoma, Chemical, Chromatography, Citric Acid Cycle, Classification, Cluster Analysis, Comparative Study, Computational Biology, Computer-Assisted, Cystadenoma, DNA, Databases, Decision Making, Diagnosis, Differential, Drug, Drug Design, Electrostatics, Eukaryotic Cells, Factual, Feasibility Studies, Female, Gene Expression, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Heterogeneity, Genetic Markers, Hemolysins, Humans, Internet, Ion Exchange, Leukemia, Ligands, Likelihood Functions, Logistic Models, Lung Neoplasms, Lymphocytic, Lymphoma, Markov Chains, Mathematics, Messenger, Models, Molecular, Molecular Probe Techniques, Molecular Sequence Data, Nanotechnology, Neoplasm, Neoplasms, Neoplastic, Neural Networks (Computer), Non-P.H.S., Non-Small-Cell Lung, Non-U.S. Gov't, Nucleic Acid Conformation, Nucleic Acid Hybridization, Observer Variation, Oligonucleotide Array Sequence Analysis, Ovarian Neoplasms, P.H.S., Pattern Recognition, Probability, Protein Binding, Protein Conformation, Proteins, Quality Control, Quantum Theory, RNA, RNA Splicing, Receptors, Reference Values, Regression Analysis, Reproducibility of Results, Research Support, Saccharomyces cerevisiae Proteins, Sensitivity and Specificity, Sequence Analysis, Signal Processing, Software, Statistical, Stomach Neoplasms, Structural, Structure-Activity Relationship, Thermodynamics, Transcription, Tumor Markers, U.S. Gov't, 12421757

MOTIVATION: A number of algorithms and analytical models have been employed to reduce the multidimensional complexity of DNA array data and attempt to extract some meaningful interpretation of the results. These include clustering, principal components analysis, self-organizing maps, and support vector machine analysis. Each method assumes an implicit model for the data, many of which separate genes into distinct clusters defined by similar expression profiles in the samples tested. A point of concern is that many genes may be involved in a number of distinct behaviours, and should therefore be modelled to fit into as many separate clusters as detected in the multidimensional gene expression space. The analysis of gene expression data using a decomposition model that is independent of the observer involved would be highly beneficial to improve standard and reproducible classification of clinical and research samples. RESULTS: We present a variational independent component analysis (ICA) method for reducing high dimensional DNA array data to a smaller set of latent variables, each associated with a gene signature. We present the results of applying the method to data from an ovarian cancer study, revealing a number of tissue type-specific and tissue type-independent gene signatures present in varying amounts among the samples surveyed. The observer independent results of such molecular analysis of biological samples could help identify patients who would benefit from different treatment strategies. We further explore the application of the model to similar high-throughput studies.

Keywords: Acute, Algorithms, Automated, Base Pair Mismatch, Base Pairing, Base Sequence, Biological, Biosensing Techniques, Cluster Analysis, Comparative Study, Computer-Assisted, Cystadenoma, DNA, Female, Gene Expression, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Markers, Hemolysins, Humans, Leukemia, Lymphocytic, Markov Chains, Messenger, Models, Molecular Probe Techniques, Molecular Sequence Data, Nanotechnology, Neoplasm, Neoplastic, Neural Networks (Computer), Non-U.S. Gov't, Nucleic Acid Conformation, Observer Variation, Oligonucleotide Array Sequence Analysis, Ovarian Neoplasms, Pattern Recognition, Quality Control, RNA, Reference Values, Reproducibility of Results, Research Support, Sensitivity and Specificity, Signal Processing, Statistical, Stomach Neoplasms, Transcription, Tumor Markers, 12490446

The ability to grow extra nodes is a potentially useful facility for a self-organising neural network. A network that can add nodes into its map space can approximate the input space more accurately, and often more parsimoniously, than a network with predefined structure and size, such as the Self-Organising Map. In addition, a growing network can deal with dynamic input distributions. Most of the growing networks that have been proposed in the literature add new nodes to support the node that has accumulated the highest error during previous iterations or to support topological structures. This usually means that new nodes are added only when the number of iterations is an integer multiple of some pre-defined constant, A. This paper suggests a way in which the learning algorithm can add nodes whenever the network in its current state does not sufficiently match the input. In this way the network grows very quickly when new data is presented, but stops growing once the network has matched the data. This is particularly important when we consider dynamic data sets, where the distribution of inputs can change to a new regime after some time. We also demonstrate the preservation of neighbourhood relations in the data by the network. The new network is compared to an existing growing network, the Growing Neural Gas (GNG), on a artificial dataset, showing how the network deals with a change in input distribution after some time. Finally, the new network is applied to several novelty detection tasks and is compared with both the GNG and an unsupervised form of the Reduced Coulomb Energy network on a robotic inspection task and with a Support Vector Machine on two benchmark novelty detection tasks.

Keywords: Acute, Algorithms, Animals, Anion Exchange Resins, Artificial Intelligence, Automated, Base Pair Mismatch, Base Pairing, Base Sequence, Biological, Biosensing Techniques, Carcinoma, Chemical, Chromatography, Citric Acid Cycle, Classification, Cluster Analysis, Comparative Study, Computational Biology, Computer-Assisted, Cystadenoma, DNA, Databases, Decision Making, Diagnosis, Differential, Drug, Drug Design, Electrostatics, Eukaryotic Cells, Factual, Feasibility Studies, Female, Gene Expression, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Heterogeneity, Genetic Markers, Hemolysins, Humans, Internet, Ion Exchange, Leukemia, Ligands, Likelihood Functions, Logistic Models, Lung Neoplasms, Lymphocytic, Lymphoma, Markov Chains, Mathematics, Messenger, Models, Molecular, Molecular Probe Techniques, Molecular Sequence Data, Nanotechnology, Neoplasm, Neoplasms, Neoplastic, Neural Networks (Computer), Non-P.H.S., Non-Small-Cell Lung, Non-U.S. Gov't, Nucleic Acid Conformation, Nucleic Acid Hybridization, Observer Variation, Oligonucleotide Array Sequence Analysis, Ovarian Neoplasms, P.H.S., Pattern Recognition, Probability, Probability Learning, Protein Binding, Protein Conformation, Proteins, Quality Control, Quantum Theory, RNA, RNA Splicing, Receptors, Reference Values, Regression Analysis, Reproducibility of Results, Research Support, Robotics, Saccharomyces cerevisiae Proteins, Sensitivity and Specificity, Sequence Analysis, Signal Processing, Software, Statistical, Stomach Neoplasms, Structural, Structure-Activity Relationship, Thermodynamics, Transcription, Tumor Markers, U.S. Gov't, 12416693

The system-wide study of proteins presents an exciting challenge in this information-rich age of whole-genome biology. Although traditional investigations have yielded abundant information about individual proteins, they have been less successful at providing us with an integrated understanding of biological systems. The promise of proteomics is that, by studying many components simultaneously, we will learn how proteins interact with each other, as well as with non-proteinaceous molecules, to control complex processes in cells, tissues and even whole organisms. Here, I discuss the role of microarray technology in this burgeoning area.

Keywords: Forecasting; Humans; Immunoassay, methods; Protein Array Analysis, methods; Proteomics, methods

Keywords: biosvm

Feature selection plays an important role in classification. We present a comparative study on six feature selection heuristics by applying them to two sets of data. The first set of data are gene expression profiles from Acute Lymphoblastic Leukemia (ALL) patients. The second set of data are proteomic patterns from ovarian cancer patients. Based on features chosen by these methods, error rates of several classification algorithms were obtained for analysis. Our results demonstrate the importance of feature selection in accurately classifying new samples.

Keywords: PUlearning

Genomics projects have resulted in a flood of sequence data. Functional annotation currently relies almost exclusively on inter-species sequence comparison and is restricted in cases of limited data from related species and widely divergent sequences with no known homologs. Here, we demonstrate that codon composition, a fusion of codon usage bias and amino acid composition signals, can accurately discriminate, in the absence of sequence homology information, cytoplasmic ribosomal protein genes from all other genes of known function in Saccharomyces cerevisiae, Escherichia coli and Mycobacterium tuberculosis using an implementation of support vector machines, SVMlight. Analysis of these codon composition signals is instructive in determining features that confer individuality to ribosomal protein genes. Each of the sets of positively charged, negatively charged and small hydrophobic residues, as well as codon bias, contribute to their distinctive codon composition profile. The representation of all these signals is sensitively detected, combined and augmented by the SVMs to perform an accurate classification. Of special mention is an obvious outlier, yeast gene RPL22B, highly homologous to RPL22A but employing very different codon usage, perhaps indicating a non-ribosomal function. Finally, we propose that codon composition be used in combination with other attributes in gene/protein classification by supervised machine learning algorithms.

Keywords: biosvm

MOTIVATION: The expression of genes is controlled by specific combinations of cellular variables. We applied Independent Component Analysis (ICA) to gene expression data, deriving a linear model based on hidden variables, which we term 'expression modes'. The expression of each gene is a linear function of the expression modes, where, according to the ICA model, the linear influences of different modes show a minimal statistical dependence, and their distributions deviate sharply from the normal distribution. RESULTS: Studying cell cycle-related gene expression in yeast, we found that the dominant expression modes could be related to distinct biological functions, such as phases of the cell cycle or the mating response. Analysis of human lymphocytes revealed modes that were related to characteristic differences between cell types. With both data sets, the linear influences of the dominant modes showed distributions with large tails, indicating the existence of specifically up- and downregulated target genes. The expression modes and their influences can be used to visualize the samples and genes in low-dimensional spaces. A projection to expression modes helps to highlight particular biological functions, to reduce noise, and to compress the data in a biologically sensible way.

Keywords: biosvm

Previous studies have established an important role of histone acetylation in transcriptional control by nuclear hormone receptors. With chromatin immunoprecipitation assays, we have now investigated whether histone methylation and phosphorylation are also involved in transcriptional regulation by thyroid hormone receptor (TR). We found that repression by unliganded TR is associated with a substantial increase in methylation of H3 lysine 9 (H3-K9) and a decrease in methylation of H3 lysine 4 (H3-K4), methylation of H3 arginine 17 (H3-R17), and a dual modification of phosphorylation of H3 serine 10 and acetylation of lysine 14 (pS10/acK14). On the other hand, transcriptional activation by liganded TR is coupled with a substantial decrease in both H3-K4 and H3-K9 methylation and a robust increase in H3-R17 methylation and the dual modification of pS10/acK14. Trichostatin A treatment results in not only histone hyperacetylation but also an increase in methylation of H3-K4, increase in dual modification of pS10/acK14, and reduction in methylation of H3-K9, revealing an extensive interplay between histone acetylation, methylation, and phosphorylation. In an effort to understand the underlying mechanism for an increase in H3-K9 methylation during repression by unliganded TR, we demonstrated that TR interacts in vitro with an H3-K9-specific histone methyltransferase (HMT), SUV39H1. Functional analysis indicates that SUV39H1 can facilitate repression by unliganded TR and in so doing requires its HMT activity. Together, our data uncover a novel role of H3-K9 methylation in repression by unliganded TR and provide strong evidence for the involvement of multiple distinct histone covalent modifications (acetylation, methylation, and phosphorylation) in transcriptional control by nuclear hormone receptors.

Keywords: Animals; Cell Fractionation; Gene Expression Regulation, drug effects; Genes, Reporter; Histone-Lysine N-Methyltransferase; Histones, chemistry/genetics/metabolism; Humans; Hydroxamic Acids, pharmacology; Methylation; Methyltransferases, metabolism; Oocytes, physiology; Phosphorylation; Protein Methyltransferases; Protein Synthesis Inhibitors, pharmacology; Receptors, Thyroid Hormone, metabolism; Transcription, Genetic; Xenopus laevis, physiology

Keywords: biosvm

The annotated genomes of organisms define a 'blueprint' of their possible gene products. Post-genome analyses attempt to confirm and modify the annotation and impose a sense of the spatial, temporal and developmental usage of genetic information by the organism. Here we describe a large-scale, high-accuracy (average deviation less than 0.02 Da at 1,000 Da) mass spectrometric proteome analysis of selected stages of the human malaria parasite Plasmodium falciparum. The analysis revealed 1,289 proteins of which 714 proteins were identified in asexual blood stages, 931 in gametocytes and 645 in gametes. The last two groups provide insights into the biology of the sexual stages of the parasite, and include conserved, stage-specific, secreted and membrane-associated proteins. A subset of these proteins contain domains that indicate a role in cell-cell interactions, and therefore can be evaluated as potential components of a malaria vaccine formulation. We also report a set of peptides with significant matches in the parasite genome but not in the protein set predicted by computational methods.

Keywords: plasmodium

We present a new approach to the induction of SARs based on the generation of structural fragments and support vector machines (SVMs). It is tailored for bio-chemical databases, where the examples are two-dimensional descriptions of chemical compounds. The fragment generator finds all fragments (i.e. linearly connected atoms) that satisfy user-specified constraints regarding their frequency and generality. In this paper, we are querying for fragments within a minimum and a maximum frequency in the dataset. After fragment generation, we propose to apply SVMs to the problem of inducing SARs from these fragments. We conjecture that the SVMs are particularly useful in this context, as they can deal with a large number of features. Experiments in the domains of carcinogenicity and mutagenicity prediction show that the minimum and the maximum frequency queries for fragments can be answered within a reasonable time, and that the predictive accuracy obtained using these fragments is satisfactory. However, further experiments will have to confirm that this is a viable approach to inducing SARs.

Keywords: biosvm

BACKGROUND: Comparative genomic hybridization (CGH) is a relatively new molecular cytogenetic method for detecting chromosomal imbalance. Karyotyping of human metaphases is an important step to assign each chromosome to one of 23 or 24 classes (22 autosomes and two sex chromosomes). Automatic karyotyping in CGH analysis is needed. However, conventional karyotyping approaches based on DAPI images require complex image enhancement procedures. METHODS: This paper proposes a simple feature extraction method, one that generates density profiles from original true color CGH images and uses normalized profiles as feature vectors without quantization. A classifier is developed by using support vector machine (SVM). It has good generalization ability and needs only limited training samples. RESULTS: Experiment results show that the feature extraction method of using color information in CGH images can improve greatly the classification success rate. The SVM classifier is able to acquire knowledge about human chromosomes from relatively few samples and has good generalization ability. A success rate of moe than 90% has been achieved and the time for training and testing is very short. CONCLUSIONS: The feature extraction method proposed here and the SVM-based classifier offer a promising computerized intelligent system for automatic karyotyping of CGH human chromosomes.

Keywords: cgh

Keywords: biosvm

To understand complex biological systems requires the integration of experimental and computational research ? in other words a systems biology approach. Computational biology, through pragmatic modelling and theoretical exploration, provides a powerful foundation from which to address critical scientific questions head-on. The reviews in this Insight cover many different aspects of this energetic field, although all, in one way or another, illuminate the functioning of modular circuits, including their robustness, design and manipulation. Computational systems biology addresses questions fundamental to our understanding of life, yet progress here will lead to practical innovations in medicine, drug discovery and engineering.

We present novel kernels that measure similarity of two RNA sequences, taking account of their secondary structures. Two types of kernels are presented. One is for RNA sequences with known secondary structures, the other for those without known secondary structures. The latter employs stochastic context-free grammar (SCFG) for estimating the secondary structure. We call the latter the marginalized count kernel (MCK). We show computational experiments for MCK using 74 sets of human tRNA sequence data: (i) kernel principal component analysis (PCA) for visualizing tRNA similarities, (ii) supervised classification with support vector machines (SVMs). Both types of experiment show promising results for MCKs.

Keywords: biosvm

Motivation: The enormous amount of protein sequence data uncovered by genome research has increased the demand for computer software that can automate the recognition of new proteins. We discuss the relative merits of various automated methods for recognizing G-Protein Coupled Receptors (GPCRs), a superfamily of cell membrane proteins. GPCRs are found in a wide range of organisms and are central to a cellular signalling network that regulates many basic physiological processes. They are the focus of a significant amount of current pharmaceutical research because they play a key role in many diseases. However, their tertiary structures remain largely unsolved. The methods described in this paper use only primary sequence information to make their predictions. We compare a simple nearest neighbor approach (BLAST), methods based on multiple alignments generated by a statistical profile Hidden Markov Model (HMM), and methods, including Support Vector Machines (SVMs), that transform protein sequences into fixed-length feature vectors. Results: The last is the most computationally expensive method, but our experiments show that, for those interested in annotation-quality classification, the results are worth the effort. In two-fold cross-validation experiments testing recognition of GPCR subfamilies that bind a specific ligand (such as a histamine molecule), the errors per sequence at the Minimum Error Point (MEP) were 13.7 SVMs, 17.1 SVM classification, 25.5 and 49 Kernel Nearest Neighbor (kernNN). The percentage of true positives recognized before the first false positive was 65 both SVM methods, 13 for kernNN. Availability: We have set up a web server for GPCR subfamily classification based on hierarchical multi-class SVMs at http://www.soe.ucsc.edu/research/compbio/gpcr-subclass. By scanning predicted peptides found in the human genome with the SVMtree server, we have identified a large number of genes that encode GPCRs. A list of our predictions for human GPCRs is available at http://www.soe.ucsc.edu/research/compbio/gpcr·hg/class·results. We also provide suggested subfamily classification for 18 sequences previously identified as unclassified Class A (rhodopsin-like) GPCRs in GPCRDB (Horn et al. , Nucleic Acids Res. , 26, 277?281, 1998), available at http://www.soe.ucsc.edu/research/compbio/gpcr/classA·unclassified/

Keywords: fisher-kernel sequence-classification biosvm

Patterns of DNA methylation and chromatin structure are profoundly altered in neoplasia and include genome-wide losses of, and regional gains in, DNA methylation. The recent explosion in our knowledge of how chromatin organization modulates gene transcription has further highlighted the importance of epigenetic mechanisms in the initiation and progression of human cancer. These epigenetic changes - in particular, aberrant promoter hypermethylation that is associated with inappropriate gene silencing - affect virtually every step in tumour progression. In this review, we discuss these epigenetic events and the molecular alterations that might cause them and/or underlie altered gene expression in cancer.

There is tremendous ferment in the field of epigenetics as the relationships between chromatin structure and DNA methylation patterns become clearer. Central to this activity is the realization that the 'histone code', which involves the post-translational modification of histones and which has important ramifications for chromatin structure, may be linked to the DNA cytosine methylation pattern. New discoveries have suggested that histone lysine 9 methylation is implicated in the spread of heterochromatin in Drosophila and other organisms. Very recently it has been found that histone lysine 9 methylation is also necessary for some DNA methylation in Neurospora and plants. There is therefore the possibility that these two processes are closely linked, suggesting ways in which DNA methylation patterns may be established during normal development. Understanding these processes is fundamental to understanding what goes awry during the process of aging and carcinogenesis where DNA methylation patterns become substantially altered and contribute to the malignant phenotype.

Keywords: Animals; Chromatin; CpG Islands; DNA Methylation; DNA, Neoplasm; Gene Expression Regulation; Gene Silencing; Histone-Lysine N-Methyltransferase; Humans; Neoplasms; Plants; Transcription, Genetic

We investigate the relationship of protein-protein interactions with mRNA expression levels, by integrating a variety of data sources for yeast. We focus on known protein complexes that have clearly defined interactions between their subunits. We find that subunits of the same protein complex show significant coexpression, both in terms of similarities of absolute mRNA levels and expression profiles, e.g., we can often see subunits of a complex having correlated patterns of expression over a time course. We classify the yeast protein complexes as either permanent or transient, with permanent ones being maintained through most cellular conditions. We find that, generally, permanent complexes, such as the ribosome and proteasome, have a particularly strong relationship with expression, while transient ones do not. However, we note that several transient complexes, such as the RNA polymerase II holoenzyme and the replication complex, can be subdivided into smaller permanent ones, which do have a strong relationship to gene expression. We also investigated the interactions in aggregated, genome-wide data sets, such as the comprehensive yeast two-hybrid experiments, and found them to have only a weak relationship with gene expression, similar to that of transient complexes. (Further details on genecensus.org/expression/interactions and bioinfo.mbb.yale.edu/expression/interactions.)

We discuss the changing use of epigenetics, a term coined by Conrad Waddington in the 1940s, and how the epigenetic approach to development differs from the genetic approach. Originally, epigenetics referred to the study of the way genes and their products bring the phenotype into being. Today, it is primarily concerned with the mechanisms through which cells become committed to a particular form or function and through which that functional or structural state is then transmitted in cell lineages. We argue that modern epigenetics is important not only because it has practical significance for medicine, agriculture, and species conservation, but also because it has implications for the way in which we should view heredity and evolution. In particular, recognizing that there are epigenetic inheritance systems through which non-DNA variations can be transmitted in cell and organismal lineages broadens the concept of heredity and challenges the widely accepted gene-centered neo-Darwinian version of Darwinism.

Keywords: csbcbook

We propose a new statistical method for constructing genetic network from microarray gene expression data by using a Bayesian network. An essential point of Bayesian network construction is in the estimation of the conditional distribution of each random variable. We consider fitting nonparametric regression models with heterogeneous error variances to the microarray gene expression data to capture the nonlinear structures between genes. A problem still remains to be solved in selecting an optimal graph, which gives the best representation of the system among genes. We theoretically derive a new graph selection criterion from Bayes approach in general situations. The proposed method includes previous methods based on Bayesian networks. We demonstrate the effectiveness of the proposed method through the analysis of Saccharomyces cerevisiae gene expression data newly obtained by disrupting 100 genes.

Keywords: biogm

We propose a new method for constructing genetic network from gene expression data by using Bayesian networks. We use nonparametric regression for capturing nonlinear relationships between genes and derive a new criterion for choosing the network in general situations. In a theoretical sense, our proposed theory and methodology include previous methods based on Bayes approach. We applied the proposed method to the S. cerevisiae cell cycle data and showed the effectiveness of our method by comparing with previous methods.

Keywords: biogm

MOTIVATION: In model organisms such as yeast, large databases of protein-protein and protein-DNA interactions have become an extremely important resource for the study of protein function, evolution, and gene regulatory dynamics. In this paper we demonstrate that by integrating these interactions with widely-available mRNA expression data, it is possible to generate concrete hypotheses for the underlying mechanisms governing the observed changes in gene expression. To perform this integration systematically and at large scale, we introduce an approach for screening a molecular interaction network to identify active subnetworks, i.e., connected regions of the network that show significant changes in expression over particular subsets of conditions. The method we present here combines a rigorous statistical measure for scoring subnetworks with a search algorithm for identifying subnetworks with high score. RESULTS: We evaluated our procedure on a small network of 332 genes and 362 interactions and a large network of 4160 genes containing all 7462 protein-protein and protein-DNA interactions in the yeast public databases. In the case of the small network, we identified five significant subnetworks that covered 41 out of 77 (53%) of all significant changes in expression. Both network analyses returned several top-scoring subnetworks with good correspondence to known regulatory mechanisms in the literature. These results demonstrate how large-scale genomic approaches may be used to uncover signalling and regulatory pathways in a systematic, integrative fashion.

Genetic changes underlie tumor progression and may lead to cancer-specific expression of critical genes. Over 1100 publications have described the use of comparative genomic hybridization (CGH) to analyze the pattern of copy number alterations in cancer, but very few of the genes affected are known. Here, we performed high-resolution CGH analysis on cDNA microarrays in breast cancer and directly compared copy number and mRNA expression levels of 13,824 genes to quantitate the impact of genomic changes on gene expression. We identified and mapped the boundaries of 24 independent amplicons, ranging in size from 0.2 to 12 Mb. Throughout the genome, both high- and low-level copy number changes had a substantial impact on gene expression, with 44% of the highly amplified genes showing overexpression and 10.5% of the highly overexpressed genes being amplified. Statistical analysis with random permutation tests identified 270 genes whose expression levels across 14 samples were systematically attributable to gene amplification. These included most previously described amplified genes in breast cancer and many novel targets for genomic alterations, including the HOXB7 gene, the presence of which in a novel amplicon at 17q21.3 was validated in 10.2% of primary breast cancers and associated with poor patient prognosis. In conclusion, CGH on cDNA microarrays revealed hundreds of novel genes whose overexpression is attributable to gene amplification. These genes may provide insights to the clonal evolution and progression of breast cancer and highlight promising therapeutic targets.

An assessment of the number of molecular targets that represent an opportunity for therapeutic intervention is crucial to the development of post-genomic research strategies within the pharmaceutical industry. Now that we know the size of the human genome, it is interesting to consider just how many molecular targets this opportunity represents. We start from the position that we understand the properties that are required for a good drug, and therefore must be able to understand what makes a good drug target.

Keywords: chemogenomics

Chemically synthesised 21-23 bp double-stranded short interfering RNAs (siRNA) can induce sequence-specific post-transcriptional gene silencing, in a process termed RNA interference (RNAi). In the present study, several siRNAs synthesised against different sites on the same target mRNA (human Tissue Factor) demonstrated striking differences in silencing efficiency. Only a few of the siRNAs resulted in a significant reduction in expression, suggesting that accessible siRNA target sites may be rare in some human mRNAs. Blocking of the 3'-OH with FITC did not reduce the effect on target mRNA. Mutations in the siRNAs relative to target mRNA sequence gradually reduced, but did not abolish mRNA depletion. Inactive siRNAs competed reversibly with active siRNAs in a sequence-independent manner. Several lines of evidence suggest the existence of a near equilibrium kinetic balance between mRNA production and siRNA-mediated mRNA depletion. The silencing effect was transient, with the level of mRNA recovering fully within 4-5 days, suggesting absence of a propagative system for RNAi in humans. Finally, we observed 3' mRNA cleavage fragments resulting from the action of the most effective siRNAs. The depletion rate-dependent appearance of these fragments argues for the existence of a two-step mRNA degradation mechanism.

Keywords: sirna

The recent abundance of genome sequence data has brought an urgent need for systematic proteomics to decipher the encoded protein networks that dictate cellular function. To date, generation of large-scale protein-protein interaction maps has relied on the yeast two-hybrid system, which detects binary interactions through activation of reporter gene expression. With the advent of ultrasensitive mass spectrometric protein identification methods, it is feasible to identify directly protein complexes on a proteome-wide scale. Here we report, using the budding yeast Saccharomyces cerevisiae as a test case, an example of this approach, which we term high-throughput mass spectrometric protein complex identification (HMS-PCI). Beginning with 10% of predicted yeast proteins as baits, we detected 3,617 associated proteins covering 25% of the yeast proteome. Numerous protein complexes were identified, including many new interactions in various signalling pathways and in the DNA damage response. Comparison of the HMS-PCI data set with interactions reported in the literature revealed an average threefold higher success rate in detection of known complexes compared with large-scale two-hybrid studies. Given the high degree of connectivity observed in this study, even partial HMS-PCI coverage of complex proteomes, including that of humans, should allow comprehensive identification of cellular networks.

Keywords: Affinity Labels, Amino Acid Sequence, Animals, Cell Cycle Proteins, Cloning, Comparative Study, DNA, DNA Damage, DNA Repair, Electrospray Ionization, Fungal, Genetic, Humans, Macromolecular Substances, Mass, Mitosis, Molecular, Molecular Sequence Data, Non-P.H.S., Non-U.S. Gov't, P.H.S., Phosphoric Monoester Hydrolases, Protein Binding, Protein Interaction Mapping, Protein Kinases, Proteome, Proteomics, Research Support, Ribonucleoproteins, Ribosomes, Saccharomyces cerevisiae, Saccharomyces cerevisiae Proteins, Sequence Alignment, Signal Transduction, Spectrometry, Spectrum Analysis, Transcription, U.S. Gov't, 11805813

The docking field has come of age. The time is ripe to present the principles of docking, reviewing the current state of the field. Two reasons are largely responsible for the maturity of the computational docking area. First, the early optimism that the very presence of the "correct" native conformation within the list of predicted docked conformations signals a near solution to the docking problem, has been replaced by the stark realization of the extreme difficulty of the next scoring/ranking step. Second, in the last couple of years more realistic approaches to handling molecular flexibility in docking schemes have emerged. As in folding, these derive from concepts abstracted from statistical mechanics, namely, populations. Docking and folding are interrelated. From the purely physical standpoint, binding and folding are analogous processes, with similar underlying principles. Computationally, the tools developed for docking will be tremendously useful for folding. For large, multidomain proteins, domain docking is probably the only rational way, mimicking the hierarchical nature of protein folding. The complexity of the problem is huge. Here we divide the computational docking problem into its two separate components. As in folding, solving the docking problem involves efficient search (and matching) algorithms, which cover the relevant conformational space, and selective scoring functions, which are both efficient and effectively discriminate between native and non-native solutions. It is universally recognized that docking of drugs is immensely important. However, protein-protein docking is equally so, relating to recognition, cellular pathways, and macromolecular assemblies. Proteins function when they are bound to other molecules. Consequently, we present the review from both the computational and the biological points of view. Although large, it covers only partially the extensive body of literature, relating to small (drug) and to large protein-protein molecule docking, to rigid and to flexible. Unfortunately, when reviewing these, a major difficulty in assessing the results is the non-uniformity in the formats in which they are presented in the literature. Consequently, we further propose a way to rectify it here.

Keywords: chemoinformatics

DNA micro-arrays now permit scientists to screen thousands of genes simultaneously and determine whether those genes are active, hyperactive or silent in normal or cancerous tissue. Because these new micro-array devices generate bewildering amounts of raw data, new analytical methods must be developed to sort out whether cancer tissues have distinctive signatures of gene expression over normal tissues or other types of cancer tissues. In this paper, we address the problem of selection of a small subset of genes from broad patterns of gene expression data, recorded on DNA micro-arrays. Using available training examples from cancer and normal patients, we build a classifier suitable for genetic diagnosis, as well as drug discovery. Previous attempts to address this problem select genes with correlation techniques. We propose a new method of gene selection utilizing Support Vector Machine methods based on Recursive Feature Elimination (RFE). We demonstrate experimentally that the genes selected by our techniques yield better classification performance and are biologically relevant to cancer. In contrast with the baseline method, our method eliminates gene redundancy automatically and yields better and more compact gene subsets. In patients with leukemia our method discovered 2 genes that yield zero leave-one-out error, while 64 genes are necessary for the baseline method to get the best result (one leave-one-out error). In the colon cancer database, using only 4 genes our method is 98% accurate, while the baseline method is only 86% accurate.

Keywords: biosvm

The idea of performing model combination, instead of model selection, has a long theoretical background in statistics. However, making use of theoretical results is ordinarily subject to the satisfaction of strong hypotheses (weak error correlation, availability of large training sets, possibility to rerun the training procedure an arbitrary number of times, etc.). In contrast, the practitioner is frequently faced with the problem of combining a given set of pre-trained classifiers, with highly correlated errors, using only a small training sample. Overfitting is then the main risk, which cannot be overcome but with a strict complexity control of the combiner selected. This suggests that SVMs should be well suited for these difficult situations. Investigating this idea, we introduce a family of multi-class SVMs and assess them as ensemble methods on a real-world problem. This task, protein secondary structure prediction, is an open problem in biocomputing for which model combination appears to be an issue of central importance. Experimental evidence highlights the gain in quality resulting from combining some of the most widely used prediction methods with our SVMs rather than with the ensemble methods traditionally used in the field. The gain increases when the outputs of the combiners are post-processed with a DP algorithm.

Keywords: biosvm

PURPOSE: To determine which machine learning classifier learns best to interpret standard automated perimetry (SAP) and to compare the best of the machine classifiers with the global indices of STATPAC 2 and with experts in glaucoma. METHODS: Multilayer perceptrons (MLP), support vector machines (SVM), mixture of Gaussian (MoG), and mixture of generalized Gaussian (MGG) classifiers were trained and tested by cross validation on the numerical plot of absolute sensitivity plus age of 189 normal eyes and 156 glaucomatous eyes, designated as such by the appearance of the optic nerve. The authors compared performance of these classifiers with the global indices of STATPAC, using the area under the ROC curve. Two human experts were judged against the machine classifiers and the global indices by plotting their sensitivity-specificity pairs. RESULTS: MoG had the greatest area under the ROC curve of the machine classifiers. Pattern SD (PSD) and corrected PSD (CPSD) had the largest areas under the curve of the global indices. MoG had significantly greater ROC area than PSD and CPSD. Human experts were not better at classifying visual fields than the machine classifiers or the global indices. CONCLUSIONS: MoG, using the entire visual field and age for input, interpreted SAP better than the global indices of STATPAC. Machine classifiers may augment the global indices of STATPAC.

Most cellular processes are carried out by multiprotein complexes. The identification and analysis of their components provides insight into how the ensemble of expressed proteins (proteome) is organized into functional units. We used tandem-affinity purification (TAP) and mass spectrometry in a large-scale approach to characterize multiprotein complexes in Saccharomyces cerevisiae. We processed 1,739 genes, including 1,143 human orthologues of relevance to human biology, and purified 589 protein assemblies. Bioinformatic analysis of these assemblies defined 232 distinct multiprotein complexes and proposed new cellular roles for 344 proteins, including 231 proteins with no previous functional annotation. Comparison of yeast and human complexes showed that conservation across species extends from single proteins to their molecular environment. Our analysis provides an outline of the eukaryotic proteome as a network of protein complexes at a level of organization beyond binary interactions. This higher-order map contains fundamental biological information and offers the context for a more reasoned and informed approach to drug discovery.

Keywords: Affinity, Affinity Labels, Amino Acid Sequence, Animals, Cell Cycle Proteins, Cells, Chromatography, Cloning, Comparative Study, Cultured, DNA, DNA Damage, DNA Repair, Electrospray Ionization, Fungal, Gene Targeting, Genetic, Humans, Macromolecular Substances, Mass, Matrix-Assisted Laser Desorption-Ionization, Mitosis, Molecular, Molecular Sequence Data, Non-P.H.S., Non-U.S. Gov't, P.H.S., Phosphoric Monoester Hydrolases, Protein Binding, Protein Interaction Mapping, Protein Kinases, Proteome, Proteomics, Recombinant Fusion Proteins, Research Support, Ribonucleoproteins, Ribosomes, Saccharomyces cerevisiae, Saccharomyces cerevisiae Proteins, Sensitivity and Specificity, Sequence Alignment, Signal Transduction, Species Specificity, Spectrometry, Spectrum Analysis, Transcription, U.S. Gov't, 11805813

The parasite Plasmodium falciparum is responsible for hundreds of millions of cases of malaria, and kills more than one million African children annually. Here we report an analysis of the genome sequence of P. falciparum clone 3D7. The 23-megabase nuclear genome consists of 14 chromosomes, encodes about 5,300 genes, and is the most (A + T)-rich genome sequenced to date. Genes involved in antigenic variation are concentrated in the subtelomeric regions of the chromosomes. Compared to the genomes of free-living eukaryotic microbes, the genome of this intracellular parasite encodes fewer enzymes and transporters, but a large proportion of genes are devoted to immune evasion and host-parasite interactions. Many nuclear-encoded proteins are targeted to the apicoplast, an organelle involved in fatty-acid and isoprenoid metabolism. The genome sequence provides the foundation for future studies of this organism, and is being exploited in the search for new drugs and vaccines to fight malaria.

Keywords: plasmodium

Sixteen dedifferentiated and pleomorphic liposarcomas were analyzed by comparative genomic hybridization (CGH) to genomic microarrays (matrix-CGH), cDNA-derived microarrays for expression profiling, and by quantitative PCR. Matrix-CGH revealed copy number gains of numerous oncogenes, i.e., CCND1, MDM2, GLI, CDK4, MYB, ESR1, and AIB1, several of which correlate with a high level of transcripts from the respective gene. In addition, a number of genes were found differentially expressed in dedifferentiated and pleomorphic liposarcomas. Application of dedicated clustering algorithms revealed that both tumor subtypes are clearly separated by the genomic profiles but only with a lesser power by the expression profiles. Using a support vector machine, a subset of five clones was identified as "class discriminators." Thus, for the distinction of these types of liposarcomas, genomic profiling appears to be more advantageous than RNA expression analysis.

Keywords: biosvm, cgh

MOTIVATION: A method for prediction of disease relevant human genes from the phenotypic appearance of a query disease is presented. Diseases of known genetic origin are clustered according to their phenotypic similarity. Each cluster entry consists of a disease and its underlying disease gene. Potential disease genes from the human genome are scored by their functional similarity to known disease genes in these clusters, which are phenotypically similar to the query disease. RESULTS: For assessment of the approach, a leave-one-out cross-validation of 878 diseases from the OMIM database, using 10672 candidate genes from the human genome, is performed. Depending on the applied parameters, in roughly one-third of cases the true solution is contained within the top scoring 3% of predictions and in two-third of cases the true solution is contained within the top scoring 15% of predictions. The prediction results can either be used to identify target genes, when searching for a mutation in monogenic diseases or for selection of loci in genotyping experiments in genetically complex diseases.

The completion of the Plasmodium falciparum clone 3D7 genome provides a basis on which to conduct comparative proteomics studies of this human pathogen. Here, we applied a high-throughput proteomics approach to identify new potential drug and vaccine targets and to better understand the biology of this complex protozoan parasite. We characterized four stages of the parasite life cycle (sporozoites, merozoites, trophozoites and gametocytes) by multidimensional protein identification technology. Functional profiling of over 2,400 proteins agreed with the physiology of each stage. Unexpectedly, the antigenically variant proteins of var and rif genes, defined as molecules on the surface of infected erythrocytes, were also largely expressed in sporozoites. The detection of chromosomal clusters encoding co-expressed proteins suggested a potential mechanism for controlling gene expression.

Keywords: plasmodium

In this paper, we investigate an approach based on support vector machines (SVMs) for detection of microcalcification (MC) clusters in digital mammograms, and propose a successive enhancement learning scheme for improved performance. SVM is a machine-learning method, based on the principle of structural risk minimization, which performs well when applied to data outside the training set. We formulate MC detection as a supervised-learning problem and apply SVM to develop the detection algorithm. We use the SVM to detect at each location in the image whether an MC is present or not. We tested the proposed method using a database of 76 clinical mammograms containing 1120 MCs. We use free-response receiver operating characteristic curves to evaluate detection performance, and compare the proposed algorithm with several existing methods. In our experiments, the proposed SVM framework outperformed all the other methods tested. In particular, a sensitivity as high as 94% was achieved by the SVM method at an error rate of one false-positive cluster per image. The ability of SVM to out perform several well-known methods developed for the widely studied problem of MC detection suggests that SVM is a promising technique for object detection in a medical imaging application.

With the continual pressure to ensure follow-up molecules to billion dollar blockbuster drugs, there is a hurdle in profitability and growth for pharmaceutical companies in the next decades. With each success and failure we increasingly appreciate that a key to the success of synthesized molecules through the research and development process is the possession of drug-like properties. These properties include an adequate bioactivity as well as adequate solubility, an ability to cross critical membranes (intestinal and sometimes blood-brain barrier), reasonable metabolic stability and of course safety in humans. Dependent on the therapeutic area being investigated it might also be desirable to avoid certain enzymes or transporters to circumvent potential drug-drug interactions. It may also be important to limit the induction of these same proteins that can result in further toxicities. We have clearly moved the assessment of in vitro absorption, distribution, metabolism, excretion and toxicity (ADME/TOX) parameters much earlier in the discovery organization than a decade ago with the inclusion of higher throughput systems. We are also now faced with huge amounts of ADME/TOX data for each molecule that need interpretation and also provide a valuable resource for generating predictive computational models for future drug discovery. The present review aims to show what tools exist today for visualizing and modeling ADME/TOX data, what tools need to be developed, and how both the present and future tools are valuable for virtual filtering using ADME/TOX and bioactivity properties in parallel as a viable addition to present practices.

Keywords: ATP-Binding Cassette Transporters, Algorithms, Animals, Biological, Biological Availability, Computer Simulation, Drug Design, Drug Evaluation, Drug Industry, Gene Expression Profiling, Humans, Models, Organic Anion Transporters, P.H.S., Pharmaceutical, Pharmaceutical Preparations, Pharmacogenetics, Pharmacokinetics, Preclinical, Proteomics, Research Support, Software, Systems Biology, Technology, Toxicity Tests, U.S. Gov't, 12489686

Background T-cells are key players in regulating a specific immune response. Activation of cytotoxic T-cells requires recognition of specific peptides bound to Major Histocompatibility Complex (MHC) class I molecules. MHC-peptide complexes are potential tools for diagnosis and treatment of pathogens and cancer, as well as for the development of peptide vaccines. Only one in 100 to 200 potential binders actually binds to a certain MHC molecule, therefore a good prediction method for MHC class I binding peptides can reduce the number of candidate binders that need to be synthesized and tested. Results Here, we present a novel approach, SVMHC, based on support vector machines to predict the binding of peptides to MHC class I molecules. This method seems to perform slightly better than two profile based methods, SYFPEITHI and HLA_BIND. The implementation of SVMHC is quite simple and does not involve any manual steps, therefore as more data become available it is trivial to provide prediction for more MHC types. SVMHC currently contains prediction for 26 MHC class I types from the MHCPEP database or alternatively 6 MHC class I types from the higher quality SYFPEITHI database. The prediction models for these MHC types are implemented in a public web service available at http://www.sbc.su.se/svmhc/. Conclusions Prediction of MHC class I binding peptides using Support Vector Machines, shows high performance and is easy to apply to a large number of MHC class I types. As more peptide data are put into MHC databases, SVMHC can easily be updated to give prediction for additional MHC class I types. We suggest that the number of binding peptides needed for SVM training is at least 20 sequences.

Keywords: biosvm immunoinformatics

The DNA of eukaryotes is wrapped around nucleosomes and packaged into chromatin. Covalent modifications of the histone proteins that comprise the nucleosome alter chromatin structure and have major effects on gene expression. Methylation of lysine 4 of histone H3 by COMPASS is required for silencing of genes located near chromosome telomeres and within the rDNA (Krogan, N. J, Dover, J., Khorrami, S., Greenblatt, J. F., Schneider, J., Johnston, M., and Shilatifard, A. (2002) J. Biol. Chem. 277, 10753-10755; Briggs, S. D., Bryk, M., Strahl, B. D., Cheung, W. L., Davie, J. K., Dent, S. Y., Winston, F., and Allis, C. D. (2001) Genes. Dev. 15, 3286-3295). To learn about the mechanism of histone methylation, we surveyed the genome of the yeast Saccharomyces cerevisiae for genes necessary for this process. By analyzing approximately 4800 mutant strains, each deleted for a different non-essential gene, we discovered that the ubiquitin-conjugating enzyme Rad6 is required for methylation of lysine 4 of histone H3. Ubiquitination of histone H2B on lysine 123 is the signal for the methylation of histone H3, which leads to silencing of genes located near telomeres.

Keywords: DNA, Ribosomal, metabolism; Electrophoresis, Polyacrylamide Gel; Gene Silencing; Histones, metabolism; Ligases, metabolism; Lysine, metabolism; Methylation; Models, Biological; Mutation; Saccharomyces cerevisiae Proteins; Saccharomyces cerevisiae, genetics; Ubiquitin, metabolism; Ubiquitin-Conjugating Enzymes

Two different machine-learning algorithms have been used to predict the blood-brain barrier permeability of different classes of molecules, to develop a method to predict the ability of drug compounds to penetrate the CNS. The first algorithm is based on a multilayer perceptron neural network and the second algorithm uses a support vector machine. Both algorithms are trained on an identical data set consisting of 179 CNS active molecules and 145 CNS inactive molecules. The training parameters include molecular weight, lipophilicity, hydrogen bonding, and other variables that govern the ability of a molecule to diffuse through a membrane. The results show that the support vector machine outperforms the neural network. Based on over 30 different validation sets, the SVM can predict up to 96 molecules correctly, averaging 81.5 of equal numbers of CNS positive and negative molecules. This is quite favorable when compared with the neural network's average performance of 75.7 the SVM algorithm are very encouraging and suggest that a classification tool like this one will prove to be a valuable prediction approach.

Keywords: biosvm

Keywords: conditional-random-field

In recent years we have witnessed an exponential increase in the amount of biological information, either DNA or protein sequences, that has become available in public databases. This has been followed by an increased interest in developing computational techniques to automatically classify these large volumes of sequence data into various categories corresponding to either their role in the chromosomes, their structure, and/or their function. In this paper we evaluate some of the widely-used sequence classification algorithms and develop a framework for modeling sequences in a fashion so that traditional machine learning algorithms, such as support vector machines, can be applied easily. Our detailed experimental evaluation shows that the SVM-based approaches are able to achieve higher classification accuracy compared to the more traditional sequence classification algorithms such as Markov model based techniques and K-nearest neighbor based approaches.

Keywords: biosvm

We describe an approach to detect the frequency of interaction between any two genomic loci. Generation of a matrix of interaction frequencies between sites on the same or different chromosomes reveals their relative spatial disposition and provides information about the physical properties of the chromatin fiber. This methodology can be applied to the spatial organization of entire genomes in organisms from bacteria to human. Using the yeast Saccharomyces cerevisiae, we could confirm known qualitative features of chromosome organization within the nucleus and dynamic changes in that organization during meiosis. We also analyzed yeast chromosome III at the G1 stage of the cell cycle. We found that chromatin is highly flexible throughout. Furthermore, functionally distinct AT- and GC-rich domains were found to exhibit different conformations, and a population-average 3D model of chromosome III could be determined. Chromosome III emerges as a contorted ring.

Keywords: AT Rich Sequence; Cell Fractionation; Cell Nucleus; Centromere; Chromatin; Chromosomes, Fungal; Cross-Linking Reagents; Deoxyribonuclease EcoRI; Formaldehyde; G1 Phase; GC Rich Sequence; Genome, Fungal; Mathematics; Meiosis; Mitosis; Polymerase Chain Reaction; Protein Conformation; Saccharomyces cerevisiae; Telomere

Motivation: The large amount of available annotated Arabidopsis thaliana sequences allows the induction of splice site prediction models with supervised learning algorithms (see Haussler (1998) for a review and references). These algorithms need information sources or features from which the models can be computed. For splice site prediction, the features we consider in this study are the presence or absence of certain nucleotides in close proximity to the splice site. Since it is not known how many and which nucleotides are relevant for splice site prediction, the set of features is chosen large enough such that the probability that all relevant information sources are in the set is very high. Using only those features that are relevant for constructing a splice site prediction system might improve the system and might also provide us with useful biological knowledge. Using fewer features will of course also improve the prediction speed of the system. Results: A wrapper-based feature subset selection algorithm using a support vector machine or a naive Bayes prediction method was evaluated against the traditional method for selecting features relevant for splice site prediction. Our results show that this wrapper approach selects features that improve the performance against the use of all features and against the use of the features selected by the traditional method. Availability: The data and additional interactive graphs on the selected feature subsets are available at http://www.psb.rug.ac.be/gps Contact: svgro@gengenp.rug.ac.be yvdp@gengenp.rug.ac.be

Keywords: biosvm

BACKGROUND: T-cells are key players in regulating a specific immune response. Activation of cytotoxic T-cells requires recognition of specific peptides bound to Major Histocompatibility Complex (MHC) class I molecules. MHC-peptide complexes are potential tools for diagnosis and treatment of pathogens and cancer, as well as for the development of peptide vaccines. Only one in 100 to 200 potential binders actually binds to a certain MHC molecule, therefore a good prediction method for MHC class I binding peptides can reduce the number of candidate binders that need to be synthesized and tested. RESULTS: Here, we present a novel approach, SVMHC, based on support vector machines to predict the binding of peptides to MHC class I molecules. This method seems to perform slightly better than two profile based methods, SYFPEITHI and HLA_BIND. The implementation of SVMHC is quite simple and does not involve any manual steps, therefore as more data become available it is trivial to provide prediction for more MHC types. SVMHC currently contains prediction for 26 MHC class I types from the MHCPEP database or alternatively 6 MHC class I types from the higher quality SYFPEITHI database. The prediction models for these MHC types are implemented in a public web service available at http://www.sbc.su.se/svmhc/. CONCLUSIONS: Prediction of MHC class I binding peptides using Support Vector Machines, shows high performance and is easy to apply to a large number of MHC class I types. As more peptide data are put into MHC databases, SVMHC can easily be updated to give prediction for additional MHC class I types. We suggest that the number of binding peptides needed for SVM training is at least 20 sequences.

Keywords: Animals; Artificial Intelligence; Comparative Study; Computational Biology; Databases, Protein; Epitopes, T-Lymphocyte; HLA Antigens; Histocompatibility Antigens Class I; Humans; Peptides; Predictive Value of Tests; Protein Binding; Research Support, Non-U.S. Gov't; Sensitivity and Specificity

Microarray technology is now widely available and is being applied to address increasingly complex scientific questions. Consequently, there is a greater demand for statistical assessment of the conclusions drawn from microarray experiments. This review discusses fundamental issues of how to design an experiment to ensure that the resulting data are amenable to statistical analysis. The discussion focuses on two-color spotted cDNA microarrays, but many of the same issues apply to single-color gene-expression assays as well.

Keywords: Animals; DNA, Complementary, analysis; Gene Expression; Gene Expression Profiling, methods; Mice; Models, Biological; Oligonucleotide Array Sequence Analysis, methods; Reference Standards; Reproducibility of Results; Research Design; Statistics as Topic

Proteins are generally classified into the following 12 subcellular locations: 1) chloroplast, 2) cytoplasm, 3) cytoskeleton, 4) endoplasmic reticulum, 5) extracellular, 6) Golgi apparatus, 7) lysosome, 8) mitochondria, 9) nucleus, 10) peroxisome, 11) plasma membrane, and 12) vacuole. Because the function of a protein is closely correlated with its subcellular location, with the rapid increase in new protein sequences entering into databanks, it is vitally important for both basic research and pharmaceutical industry to establish a high throughput tool for predicting protein subcellular location. In this paper, a new concept, the so-called "functional domain composition" is introduced. Based on the novel concept, the representation for a protein can be defined as a vector in a high-dimensional space, where each of the clustered functional domains derived from the protein universe serves as a vector base. With such a novel representation for a protein, the support vector machine (SVM) algorithm is introduced for predicting protein subcellular location. High success rates are obtained by the self-consistency test, jackknife test, and independent dataset test, respectively. The current approach not only can play an important complementary role to the powerful covariant discriminant algorithm based on the pseudo amino acid composition representation (Chou, K. C. (2001) Proteins Struct. Funct. Genet. 43, 246-255; Correction (2001) Proteins Struct. Funct. Genet. 44, 60), but also may greatly stimulate the development of this area.

Keywords: biosvm

Hepatocellular carcinoma (HCC) is a leading cause of death worldwide. Using cDNA microarrays to characterize patterns of gene expression in HCC, we found consistent differences between the expression patterns in HCC compared with those seen in nontumor liver tissues. The expression patterns in HCC were also readily distinguished from those associated with tumors metastatic to liver. The global gene expression patterns intrinsic to each tumor were sufficiently distinctive that multiple tumor nodules from the same patient could usually be recognized and distinguished from all the others in the large sample set on the basis of their gene expression patterns alone. The distinctive gene expression patterns are characteristic of the tumors and not the patient; the expression programs seen in clonally independent tumor nodules in the same patient were no more similar than those in tumors from different patients. Moreover, clonally related tumor masses that showed distinct expression profiles were also distinguished by genotypic differences. Some features of the gene expression patterns were associated with specific phenotypic and genotypic characteristics of the tumors, including growth rate, vascular invasion, and p53 overexpression.

Keywords: csbcbook-ch3, csbcbook

Glaucoma is a progressive optic neuropathy with characteristic structural changes in the optic nerve head reflected in the visual field. The visual-field sensitivity test is commonly used in a clinical setting to evaluate glaucoma. Standard automated perimetry (SAP) is a common computerized visual-field test whose output is amenable to machine learning. We compared the performance of a number of machine learning algorithms with STATPAC indexes mean deviation, pattern standard deviation, and corrected pattern standard deviation. The machine learning algorithms studied included multilayer perceptron (MLP), support vector machine (SVM), and linear (LDA) and quadratic discriminant analysis (QDA), Parzen window, mixture of Gaussian (MOG), and mixture of generalized Gaussian (MGG). MLP and SVM are classifiers that work directly on the decision boundary and fall under the discriminative paradigm. Generative classifiers, which first model the data probability density and then perform classification via Bayes' rule, usually give deeper insight into the structure of the data space. We have applied MOG, MGG, LDA, QDA, and Parzen window to the classification of glaucoma from SAP. Performance of the various classifiers was compared by the areas under their receiver operating characteristic curves and by sensitivities (true-positive rates) at chosen specificities (true-negative rates). The machine-learning-type classifiers showed improved performance over the best indexes from STATPAC. Forward-selection and backward-elimination methodology further improved the classification rate and also has the potential to reduce testing time by diminishing the number of visual-field location measurements.

Keywords: Acute, Algorithms, Animals, Anion Exchange Resins, Artificial Intelligence, Automated, Base Pair Mismatch, Base Pairing, Base Sequence, Biological, Biosensing Techniques, Carcinoma, Chemical, Chromatography, Citric Acid Cycle, Classification, Cluster Analysis, Comparative Study, Computational Biology, Computer-Assisted, Cystadenoma, DNA, Databases, Decision Making, Diagnosis, Differential, Discriminant Analysis, Drug, Drug Design, Electrostatics, Epitopes, Eukaryotic Cells, Factual, False Negative Reactions, False Positive Reactions, Feasibility Studies, Female, Gene Expression, Gene Expression Profiling, Gene Expression Regulation, Genes, Genetic, Genetic Heterogeneity, Genetic Markers, Glaucoma, HLA Antigens, Hemolysins, Histocompatibility Antigens Class I, Humans, Internet, Intraocular Pressure, Ion Exchange, Lasers, Leukemia, Ligands, Likelihood Functions, Logistic Models, Lung Neoplasms, Lymphocytic, Lymphoma, Markov Chains, Mathematics, Messenger, Models, Molecular, Molecular Probe Techniques, Molecular Sequence Data, Nanotechnology, Neoplasm, Neoplasms, Neoplastic, Neural Networks (Computer), Neurological, Non-P.H.S., Non-Small-Cell Lung, Non-U.S. Gov't, Nucleic Acid Conformation, Nucleic Acid Hybridization, Observer Variation, Oligonucleotide Array Sequence Analysis, Open-Angle, Ophthalmoscopy, Optic Disk, Optic Nerve Diseases, Ovarian Neoplasms, P.H.S., Pattern Recognition, Peptides, Perimetry, Predictive Value of Tests, Probability, Probability Learning, Protein, Protein Binding, Protein Conformation, Proteins, Quality Control, Quantum Theory, RNA, RNA Splicing, ROC Curve, Receptors, Reference Values, Regression Analysis, Reproducibility of Results, Research Support, Robotics, Saccharomyces cerevisiae Proteins, Sensitivity and Specificity, Sequence Analysis, Signal Processing, Software, Statistical, Stomach Neoplasms, Structural, Structure-Activity Relationship, T-Lymphocyte, Thermodynamics, Transcription, Tumor Markers, U.S. Gov't, 12214886

In this paper, we present a pharmacophore for QT-prolonging drugs, along with a 3D QSAR (CoMFA) study for a series of very structurally variegate HERG K(+) channel blockers. The blockade of HERG K(+) channels is one of the most important molecular mechanisms through which QT-prolonging drugs increase cardiac action potential duration. Since QT prolongation is one of the most undesirable side effects of drugs, we first tried to identify the minimum set of molecular features responsible for this action and then we attempted to develop a quantitative model correlating the 3D stereoelectronic characteristics of the molecules with their HERG blocking potency. Having considered an initial set of 31 QT-prolonging drugs for which the HERG K(+) channel blocking activity was measured on mammalian transfected cells, we started the construction of a theoretical screening tool able to predict whether a new molecule can interact with the HERG channel and eventually induce the long QT syndrome. This in silico tool might be useful in the design of new drug candidates devoid of the physicochemical features likely to cause the above-mentioned side effect.

Keywords: chemoinformatics herg

In this paper, we apply a new machine learning method which is called support vector machine to approach the prediction of protein structural class. The support vector machine method is performed based on the database derived from SCOP which is based upon domains of known structure and the evolutionary relationships and the principles that govern their 3D structure. As a result, high rates of both self-consistency and jackknife test are obtained. This indicates that the structural class of a protein inconsiderably correlated with its amino and composition, and the support vector machine can be referred as a powerful computational tool for predicting the structural classes of proteins.

Keywords: biosvm

Support Vector Machine (SVM), which is one class of learning machines, was applied to predict the subcellular location of proteins by incorporating the quasi-sequence-order effect (Chou [2000] Biochem. Biophys. Res. Commun. 278:477-483). In this study, the proteins are classified into the following 12 groups: (1) chloroplast, (2) cytoplasm, (3) cytoskeleton, (4) endoplasmic reticulum, (5) extracellular, (6) Golgi apparatus, (7) lysosome, (8) mitochondria, (9) nucleus, (10) peroxisome, (11) plasma membrane, and (12) vacuole, which account for most organelles and subcellular compartments in an animal or plant cell. Examinations for self-consistency and jackknife testing of the SVMs method were conducted for three sets consisting of 1,911, 2,044, and 2,191 proteins. The correct rates for self-consistency and the jackknife test values achieved with these protein sets were 94 and 83 89 and 75 for correct prediction rates were undertaken with three independent testing datasets containing 2,148 proteins, 2,417 proteins, and 2,494 proteins producing values of 84, 77, and 74%, respectively.

Keywords: biosvm

The support vector machines (SVMs) method is proposed because it can reflect the sequence-coupling effect for a tetrapeptide in not only a beta-turn or non-beta-turn, but also in different types of beta-turn. The results of the model for 6022 tetrapeptides indicate that the rates of self-consistency for beta-turn types I, I', II, II', VI and VIII and non-beta-turns are 99.92 98.02 training data, the rate of correct prediction by the SVMs for a given protein: rubredoxin (54 residues. 51 tetrapeptides) which includes 12 beta-turn type I tetrapeptides, 1 beta-turn type II tetrapeptide and 38 non-beta-turns reached 82.4 of the SVMs implies that the formation of different beta-turn types or non-beta-turns is considerably correlated with the sequence of a tetrapeptide. The SVMs can save CPU time and avoid the overfitting problem compared with the neural network method.

Keywords: biosvm

Support Vector Machines (SVMs) which is one kind of learning machines, was applied to predict the specificity of GalNAc-transferase. The examination for the self-consistency and the jackknife test of the SVMs method were tested for the training dataset (305 oligopeptides), the correct rate of self-consistency and jackknife test reaches 100 and 84.9 testing dataset (30 oligopeptides) was tested, the rate reaches 76.67%.

Keywords: biosvm

Knowledge of the polyprotein cleavage sites by HIV protease will refine our understanding of its specificity, and the information thus acquired is useful for designing specific and efficient HIV protease inhibitors. The pace in searching for the proper inhibitors of HIV protease will be greatly expedited if one can find an accurate, robust, and rapid method for predicting the cleavage sites in proteins by HIV protease. In this article, a Support Vector Machine is applied to predict the cleavability of oligopeptides by proteases with multiple and extended specificity subsites. We selected HIV-1 protease as the subject of the study. Two hundred ninety-nine oligopeptides were chosen for the training set, while the other 63 oligopeptides were taken as a test set. Because of its high rate of self-consistency (299/299 = 100 and correct prediction rate (55/63 = 87 Support Vector Machine method can be referred to as a useful assistant technique for finding effective inhibitors of HIV protease, which is one of the targets in designing potential drugs against AIDS. The principle of the Support Vector Machine method can also be applied to analyzing the specificity of other multisubsite enzymes.

Keywords: biosvm