Teresa A. Webster

Algorithms for large-scale genotyping microarrays.

MOTIVATION Analysis of many thousands of single nucleotide polymorphisms (SNPs) across whole genome is crucial to efficiently map disease genes and understanding susceptibility to diseases, drug efficacy and side effects for different populations and individuals. High density oligonucleotide microarrays provide the possibility for such analysis with reasonable cost. Such analysis requires accurate, reliable methods for feature extraction, classification, statistical modeling and filtering. RESULTS We propose the modified partitioning around medoids as a classification method for relative allele signals. We use the average silhouette width, separation and other quantities as quality measures for genotyping classification. We form robust statistical models based on the classification results and use these models to make genotype calls and calculate quality measures of calls. We apply our algorithms to several different genotyping microarrays. We use reference types, informative Mendelian relationship in families, and leave-one-out cross validation to verify our results. The concordance rates with the single base extension reference types are 99.36% for the SNPs on autosomes and 99.64% for the SNPs on sex chromosomes. The concordance of the leave-one-out test is over 99.5% and is 99.9% higher for AA, AB and BB cells. We also provide a method to determine the gender of a sample based on the heterozygous call rate of SNPs on the X chromosome. See http://www.affymetrix.com for further information. The microarray data will also be available from the Affymetrix web site. AVAILABILITY The algorithms will be available commercially in the Affymetrix software package.

Compass: A shape-based machine learning tool for drug design

SummaryBuilding predictive models for iterative drug design in the absence of a known target protein structure is an important challenge. We present a novel technique, Compass, that removes a major obstacle to accurate prediction by automatically selecting conformations and alignments of molecules without the benefit of a characterized active site. The technique combines explicit representation of molecular shape with neural network learning methods to produce highly predictive models, even across chemically distinct classes of molecules. We apply the method to predicting human perception of musk odor and show how the resulting models can provide graphical guidance for chemical modifications.

ARIADNE: pattern-directed inference and hierarchical abstraction in protein structure recognition

The macro-molecular structural conformations of proteins exhibit higher order regularities whose recognition is complicated by many factors. ARIADNE searches for similarities between structural descriptors and hypothesized protein structure at levels more abstract than the primary sequence.

A modified Chou and Fasman protein structure algorithm

A FORTRAN program PRSTRC has been developed for protein secondary structure prediction, which is a modified Chou and Fasman (1978) analysis. This implementation carries out a running average of amino acid structure occurrence frequencies, utilizes a simple set of nucleation conditions, and allows user control over nucleation threshold and cutoff parameters. The algorithm includes prediction of the newly defined secondary structure elements: omega loops (1986). It also generates a charge distribution and hydropathy profile. Output includes a simple graphic display for a printer, or a CRT using color addition. Correct structures are predicted for T. dyscritum hemerythrin and the variable domain of mouse immunoglobin k-chain.

Massively parallel symbolic induction of protein structure/function relationships

Reports the development and implementation of efficient algorithms for several symbolic machine learning induction operators on a massively parallel computer. The authors invoke these operators as hardware induction subroutines under the control of a higher-level front-end LISP program. For them, the key contribution of this work is its demonstration of the scalability of the algorithms involved. The time complexity of the induction algorithms is essentially independent of the total size of the instance data pool, with essentially linear space (hardware) complexity. Everything described has been implemented in Common LISP or PARIS. The PARIS portion runs on a CM-2 Connection Machine. The system (ARIEL) has been applied to the DNA polymerases and to the transcriptional activators by domain experts.<<ETX>>