A new hidden Markov model method (SAM-T98) for nding remote homologs of protein sequences is described and evaluated. The method begins with a single target sequence and iteratively builds a hidden Markov model (hmm) from the sequence and homologs found using the hmm for database search. SAM-T98 is also used to construct model libraries automatically from sequences in structural databases. We evaluate the SAM-T98 method with four datasets. Three of the test sets are fold-recognition tests, where the correct answers are determined by structural similarity. The fourth uses a curated database. The method is compared against wu-blastp and against double-blast, a two-step method similar to ISS, but using blast instead of fasta. Results SAM-T98 had the fewest errors in all tests| dramatically so for the fold-recognition tests. At the minimum-error point on the SCOP-domains test, SAM-T98 got 880 true positives and 68 false positives, double-blast got 533 true positives with 71 false positives, and wu-blastp got 353 true positives with 24 false positives. The method is optimized to recognize superfamilies, and would require parameter adjustment to be used to nd family or fold relationships. One key to the performance of the hmm method is a new score-normalization technique that compares the score to the score with a reversed model rather than to a uniform null model. Availability A World Wide Web server, as well as information on obtaining the Sequence Alignment and PREPRINT to appear in Bioinformatics, 1999

Abstract not found

A new method for detecting remote protein homologies is introduced and shown to perform well in classifying protein domains by SCOP superfamily. The method is a variant of support vector machines using a new kernel function. The kernel function is derived from a generative statistical model for a protein family, in this case a hidden Markov model. This general approach of combining generative models like HMMs with discriminative methods such as support vector machines may have applications in other areas of biosequence analysis as well.

The sequences of related proteins can diverge beyond the point where their relationship can be recognised by pairwise sequence comparisons. In attempts to overcome this limitation, methods have been developed that use as a query, not a single sequence, but sets of related sequences or a representation of the characteristics shared by related sequences. Here we describe an assessment of three of these methods: the SAM-T98 implementation of a hidden Markov model procedure; PSI-BLAST; and the intermediate sequence search (ISS) procedure. We determined the extent to which these procedures can detect evolutionary relationships between the members of the sequence database PDBD40-J. This database, derived from the structural classification of proteins (SCOP), contains the sequences of proteins of known structure whose sequence identities with each other are 40 % or less. The evolutionary relationships that exist between those that have low sequence identities were found by the examination of their structural details and, in many cases, their functional

A new method, called the Fisher kernel method, for detecting remote protein homologies is introduced and shown to perform well in classifying protein domains by SCOP superfamily. The method is a variant of support vector machines using a new kernel function. The kernel function is derived from a hidden Markov model. The general approach of combining generative models like HMMs with discriminative methods such as support vector machines may have applications in other areas of biosequence analysis as well.

This paper presents the mathematical foundations of Dirichlet mixtures, which have been used to improve database search results for homologous sequences, when a variable number of sequences from a protein family or domain are known. We present a method for condensing the information in a protein database into a mixture of Dirichlet densities. These mixtures are designed to be combined with observed amino acid frequencies, to form estimates of expected amino acid probabilities at each position in a profile, hidden Markov model, or other statistical model. These estimates give a statistical model greater generalization capacity, such that remotely related family members can be more reliably recognized by the model. Dirichlet mixtures have been shown to outperform substitution matrices and other methods for computing these expected amino acid distributions in database search, resulting in fewer false positives and false negatives for the families tested. This paper corrects a previously p...

This thesis shows how to design and tune a hidden Markov model to extract factual information from a corpus of machine-readable English prose. In particular, the thesis presents a HMM that classifies and parses natural language assertions about genes being located at particular positions on chromosomes. The facts extracted by this HMM can be inserted into biological databases. The HMM is trained on a small set of sentence fragments chosen from the collected scientific abstracts in the OMIM (On-Line Mendelian Inheritance in Man) database and judged to contain the target binary relationship between gene names and gene locations. Given a novel sentence, all contiguous fragments are ranked by log-odds score, i.e. the log of the ratio of the probability of the fragment according to the target HMM to that according to a "null" HMM trained on all OMIM sentences. The most probable path through the HMM gives bindings for the annotations with precision as high as 80%. In contrast with traditional natural language processing methods, this stochastic approach makes no use either of part-of-speech taggers or dictionaries, instead employing non-emitting states to assemble modules roughly corresponding to noun, verb, and prepostional phrases. Algorithms for reestimating parameters for HMMs with non-emitting states are presented in detail. The ability to tolerate new words and recognize a wide variety of syntactic forms arises from the judicious use of "gap" states.

Protein structure prediction, to discover the fold and hence information about the probable function of the sequence of a gene about which nothing is known, is possible via homology to a sequence of

Motivation: Modeling families of related biological sequences using Hidden Markov models (HMMs), although increasingly widespread, faces at least one major problem: because of the complexity of these mathematical models, they require a relatively large training set in order to accurately characterize a given family. For families in which there are few known sequences, a standard linear HMM contains too many parameters to be trained adequately. Results: This work attempts to solve that problem by generating smaller HMMs which precisely model only the conserved regions of the family. These HMMs are constructed from motif models generated by the EM algorithm using the MEME software. Because motif-based HMMs have relatively few parameters, they can be trained using smaller data sets. Studies of shortchain alcohol dehydrogenases and 4Fe-4S ferredoxins support the claim that motif-based HMMs exhibit increased sensitivity and selectivity in database searches, especially when training sets co...

We discuss how methods based on hidden Markov models performed in the fold recognition section of the CASP2 experiment. Hidden Markov models were built for a set of about a thousand structures from the PDB database, and each CASP2 target sequence was scored against this library of hidden Markov models. In addition, a hidden Markov model was built for each of the target sequences, and all of the sequences in PDB were scored against that target model. Having high scores from both methods was found to be highly indicative of the target and a structure being homologous. Predictions were made based on several criteria: the scores with the structure models, the scores with the target models, consistency between the secondary structure in the known structure and predictions for the target (using the program PhD), human examination of predicted alignments between target and structure (using RASMOL), and solvation preferences in the alignment of the target and structure. The method worked well in comparison to other methods used at CASP2 for targets of moderate difficulty, where the closest structure in PDB could be aligned to the target with at least 15 % residue identity. There was no evidence for the method's e ectiveness for harder cases, where the residue identity was much lower than 15%.