A random duplication/deletion model in genome arrangement

A phylogeny is a tree that represents the common ancestry and evolutionary history of a set of species. Most inferential approaches use DNA sequence data to reconstruct phylogenetic trees. Genome arrangements are another source of genetic information for phylogenetic inference of distantly related species. Many existing methods of analysis assume that gene inversion is the only mechanism that accounts for the genome arrangements. However, the biological literature suggests that there is considerable evidence that a process of duplication followed by random deletion is the most prevalent mechanism of mitochondrial genome rearrangements. This dissertation will provide the first statistical methodology for analyzing arrangements under this mechanism of rearrangement.;We build a statistical phylogenetic model assuming that the duplication/deletion process is the only mechanism for genome rearrangement. To get the distribution of phylogenetic trees for a collection of species of interest, we consider the history of rearrangement events on edges together with the arrangements on nodes of a tree. To carry out a Bayesian analysis of phylogeny from genome arrangement data, we present an MCMC technique that considers all possible tree topologies, genome arrangements at internal nodes, and histories of rearrangement events based on a likelihood model of the duplication/deletion process for genome rearrangement and an appropriate prior distribution.