PolyPhy (Phylogeny reconstraction with Polytomy identification) is a java based tool for whole-genome gene set phylogeney inferences.

Phylogenetic tree, showing ancestral relations among organisms, dichotomy (bifurcation) or polytomy (multifurcating). NCBI taxonomy represents almost half of the braches as polytomies, current phylogenetic tree construction methods typically generate bifurcating trees only. In order to use bifurcating species tree to represent the true evolutionary relationships automatically, we developed PolyPhy, a stand-alone phylogeny construction with polytomy identification tool. PolyPhy first constructes a microbial phylogeny with calculating a composite distance based on the comparison of complete gene sets between two genomes using ComPhy. And then it identifies bifurcating polytomies using Bayesian Logistic Regression classification from it. It does not require multiple sequence alignment for genome distance calculation, and is able to apply different genome distance measures to the classification for dichotomy and polytomy. It is fully automated. The composite distance of ComPhy between two genomes is defined by three components: Gene Dispersion Distance (GDD), Gene Content Distance (GCD), and Genome Breakpoint-based Distance (GBD). The tree topological features extracted from bifurcating tree for Bayesian Logistic Regression are leaf-rate, intraR rate and interR rate. The phylogenetic tree is constructed using a neighbour joining method using Phylip from the distance matrix calculated by the composite distance. PolyPhy also provides the performance measure of the generated tree against Bergey's taxonomy system.

If you have any comments or suggestion, please contact:
Dr.Dong Xu
Digital Biology Laboratory
Computer Science Department and Christopher S. Bond Life Sciences Center
University of Missouri Columbia
Email: xudong@missouri.edu