Bayesian Inference in Phylogeny via Sequential Stochastic Approximation Monte Carlo

Abstract

The Sequential Stochastic approximation Monte Carlo(SSAMC) algorithm has recently been proposed by Cheon and Liang(2008) as a new phylogenetic tree construction method. SSAMC is an efficient algorithm to alleviate local trap problems and the curse of dimensionality in simulations simultaneously by making use of the sequential structure of phylogenetic trees in conjunction with stochastic approximation Monte Carlo(SAMC) simulations. In this paper, we discuss the application of SSAMC to the Bayesian inference in phylogeny. Two real datasets are used for SSAMC to show the capability of a phylogeny tree reconstruction and existing Bayesian methods, BAMBE and MrBayes, are applied for comparison. Numerical results indicate that SSAMC is a useful algorithm for phylogeny inference in terms of quality of the resulting phylogenetic trees.