|
Activity Number:
|
357
|
|
Type:
|
Invited
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
WNAR
|
| Abstract - #305089 |
|
Title:
|
A Bayesian Approach to Gene Tree Concordance
|
|
Author(s):
|
Bret Larget*+
|
|
Companies:
|
University of Wisconsin-Madison
|
|
Address:
|
Department of Statistics, Madison, WI, 53706,
|
|
Keywords:
|
phylogeny ; MCMC ; gene tree ; Bayesian statistics ; evolution ; phylogenetics
|
|
Abstract:
|
Bayesian phylogenetics involves the estimation of evolutionary relationships from genetic data. It is not unusual for different genes to support different evolutionary histories. The two most common strategies to deal with this are combining all of the data into a single analysis that assumes a single common tree or making separate independent estimates for each gene. We describe an approach for the estimation of several gene trees between these two extremes that accommodates the possibility of multiple different gene trees, but also incorporates information from all genes through a prior distribution on the set of gene trees for each individual gene tree estimate. We use a novel two-stage Markov chain Monte Carlo approach for calculations for this problem.
|
