A Mutual Information Based Sequence Distance For Vertebrate Phylogeny Using Complete Mitochondrial Genomes

This thesis present systematically a mutual information based sequence distance which is applied establish vertebrate phylogeny using complete mitochon-drial genomes. Traditional sequence distances require an alignment. But on whole genome phylogeny,events such as rearrangements make full length alignments impossible. Besides, no statistical models exist for the evolution of complete genomes currently. So, existing tools and methods such as multiple alignment and various sequence evolutionary models do not directly apply to complete genomes. At present, some distances without alignment have been developed for phylogenetic analysis using complete genomes. A new mutual information based sequence distance without alignment is defined in this thesis. This distance is also based on compositional vectors of DNA sequences or protein sequences from complete genomes. First we establish the mathematical foundation of this distance. Then this distance is applied to analyze the phylogenetic relationship of 64 vertebrates using complete mitochondrial genomes. The phylogenetic tree show that the mito-chondrial genomes are separated into three major groups. One group corresponds to mammals; one group corresponds to fish; and the other one is Archosauria (including birds and reptiles). The structure of the tree based on our new distance are roughly in agreement in topology with the current known phylogenies of vertebrates.