New York University
School of Medicine
October 2, 2014
Bayesian phylogenetics with BEAST
The theory of evolution is perhaps the only unifying theory in biology. The systematic effect of the theory evolution is best understood through phylogenetic reconstruction, which is a highly complicated inverse inference problem involving parameters in both discrete and continuous spaces. Phylogenetic inference based solely on point estimates, such as maximum likelihood, suffer from biases, which are best handled in the Bayesian context. In this talk, I will present the problem of Bayesian phylogenetic inference and methods for model selection. I will talk about the implementation of some of the underlying models and algorithms in a popular phylogenetic reconstruction software BEAST (Bayesian Evolutionary Analysis by Sampling Trees).