UCSC Bioinformatics Research Projects

DNA Research

A review of genefinding methods

Improved splice site detection in genie, from RECOMB97

Integrating database homology in a probabilistic gene structure model

A generalized hidden Markov model for the recognition of human genes in DNA, from ISMB96

EcoParse, a mailserver for finding genes in E. coli DNA that uses a Hidden Markov Model

Optimal parse of DNA, a general gene-finding program for prokaryotes that allows the user to add or remove "sensors" that look for coding signals

Using simple Markov models to search DNA databases

Linear hidden Markov models for protein and nucleic acid modeling (SAM)

RNA Research

Detecting base-pairs in RNA multiple alignments

Stochastic context-free grammars for RNA modeling

Gibbs sampling for creating stochastic context-free grammars for RNA modeling

Pseudoknot modeling using intersections of stochastic context-free grammars with applications to database search

Protein Research

Investigation of non-pairwise protein structure score functions using sets of decoy structures, Ph.D. dissertation by Christian Barrett (compressed postscript | compressed pdf)

Linear hidden Markov models for protein and nucleic acid modeling (SAM)

Getting the most out of hidden Markov models, a tutorial presented at ISMB99 (Intelligent Systems in Molecular Biology)

Predicting protein structure using only sequence information (Proteins: Structure, Function, and Genetics 1999; 37(Suppl 3):121-5) (official journal site), the results of SAM-T98 at CASP3

Remote homolog detection with SAM-T98 (Bioinformatics 1998;14(10):846-56), a more detailed presentation of SAM-T98 than the Proteins paper (above)

Sequence comparisons using multiple sequences detect three times as many remote homologues as pairwise methods (J Molec Biology, 1998; 284(4):249-51)

A discriminative framework for detecting remote protein homologies (Abstract | Postscript)

Weighting hidden Markov models for maximum discrimination (Bioinformatics 1998; 14(9):772-82) (Postscript is also available)

Visualization and sonification of protein structure sequence alignments

Predicting protein structure using hidden Markov model (Proteins: Structure, Function, and Genetics 1997; 29(Suppl 1):134-9), fold recognition in CASP2 (official journal site)

Hidden Markov models in computational biology: applications to protein modeling (J Molec Biology 1994;235:1501-3)
(compressed postscript of longer tech. rep. version, part 1 | second part of postscript)

Classifying G-protein coupled receptors with support vector machines, master's thesis, June 2000 (compressed postscript | pdf)

Estimating Amino Acid Distributions

Dirichlet mixtures, a method for improving the estimation of amino acid distributions

Evaluating regularizers for estimating distributions of amino acids, from ISMB95

Regularizers for estimating distributions of amino acids from small samples, by Kevin Karplus (UCSC-CRL-95-11)

Parallel Sequence Analysis

Kestrel parallel processor

SAM-T08 sequence alignment and hidden Markov modeling system

Discriminative Models in Computational Biology

Support vector machine classification and validation of cancer tissue samples using microarray expression data (Postscript of technical report)

Support vector machine classification of microarray gene expression data (UCSC-CRL-99-09) (Postscript of technical report)

Exploiting generative models in discriminative classifiers (Abstract | Postscript)

A discriminative framework for detecting remote protein homologies (Abstract | Postscript | 56mb compressed)

Classifying G-protein coupled receptors with support vector machines (Karchin R, Karplus K, Haussler D. Bioinformatics 2002; 18(1):147-59)

EST Analysis

A probabilistic approach to consensus multiple alignment (Postscript)

Towards an accurate EST consensus (Postscript)

Software Tools

LaTeX style file for CABIOS