Genome-wide microbial phylogeny reconstruction with polytomy identification

Abstract

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT REQUEST OF AUTHOR.] Evolutionary comparative genomics is one of the most studied areas in the computational biology. With ever increasing speed of newly sequenced microbial genome deposition, the study of the evolutionary relationships among microbes has becoming more and more imperative. Using complete genome sequences to infer large scale multi-furcating microbial phylogeny without relying on MSA (multiple sequence alignment) is greatly needed. In this dissertation work, we present two applications, ComPhy (composite distance phylogeny) and PolyPhy (phylogeny reconstruction with polytomy identification), which are intend to target two separate issues in microbial phylogeny. ComPhy uses a composite distance matrix calculated from whole-genome structural features derived by comparing the physical locations of the complete gene sets between genome pairs to produce a bifurcating phylogeny, avoiding possible incongruence introduced by MSA on selected genes. While PolyPhy employs a machine learning technique, BLR (Bayesian logistic regression) classifier to identify possible bifurcating subtrees as polytomies or not from the result trees generated from ComPhy. We have developed a set of two phylogenetic analysis applications, which are fast and robust for genome-wide inference of evolution relationships among microbial genomes.