Bayesian Bioinformatics
Speaker: Michael I. Jordan , Dept of EECS, Dept of Statistics, Univ of Calf, Berkeley
Date: February 14 2005
Time: 11:30AM to 1:00PM
Location: 32-G575
Host: P Clot/ BC & B Berger/ MIT
Contact: Kathleen Dickey, 617 253-3037, kvdickey@mit.edu
Relevant URL: http://www-math.mit.edu/compbiosem/
sorry for duplicate emails.....
abstract:
One of the advantages of Bayesian statistics is its ability to integrate multiple sources of information---to "share statistical strength" among components of a hierarchical probabilistic model. This virtue has particular relevance in bioinformatics, where many core inferential problems---e.g., phylogenetic
analysis, linkage analysis, haplotype phasing---are already naturally expressed in probabilistic terms. I will discuss our recent work on Bayesian inferential methodology in three areas in bioinformatics: (1) The prediction of molecular function based on homology, where Bayesian inference in a reconciled phylogeny is used to predict function of unannotated proteins from sparse functional annotations. We present results for 100 Pfam families. (2) Haplotype clustering and phasing, using a novel approach based on Dirichlet process priors. I present a hierarchical version of the Dirichlet process which allows us to infer relationships among haplotypes in multiple sub-populations. (3) Comparative genomics, where phylogenetic inference combines with Markovian inference to provide sensitive detection of conserved structure.
[Joint work with Steven Brenner, Barbara Engelhardt, Jon McAuliffe, Lior Pachter, Yee Whye Teh, and Eric Xing].
Department of Mathematics
& The Theory of
Computation Group
At CSAIL
See other events that are part of Bioinformatics Seminar Series 2004/2005
See other events happening in February 2005
