Exploiting phylogenetics to understand genome evolution in both modern and ancestral organisms

In this dissertation, computational evolutionary analyses, particularly phylogenetics and ancestral reconstruction, have been extensively exploited to better understand both functional divergence within individual gene families on the small-scale as well as gene content/organization at the genomic level on the large-scale. These small-scale studies focus on two gene families, thioredoxin and catenin, intended to deepen our understanding of both protein adaptation and innovation of new gene families through duplication events, respectively. Alternatively, the large-scale studies focus on both reassortments as revealed by diverse genotypes of H5N1 avian influenza viruses as well as inferences of gene content and genome rearrangements as revealed by ancestral genome reconstruction of a hypothetical ancient Mycoplasma species.;Such evolutionary studies provide us with insights into biological phenomena that in turn can be exploited for different purposes. For instance, studies of viral epidemics and modes of transmission by assigning genotypes of H5N1 highly pathogenic avian influenza viruses can help us to better prepare, prevent and control diseases. Determining functional divergence following an array of duplications within a cancer-related catenin gene family improves our understanding of developmental physiology within Metazoans. Resurrected ancient thioredoxin proteins based on computational ancestral sequence reconstruction provide possible clues to the environments that hosted early life. A set of genes inferred computationally to compose an ancient genome using Mycoplasma allow us to link genotypes with lineage-specific phenotypes and also facilitate synthetic biology's attempt to create a viable, self-sustainable organism consisting of a recombinant, minimal genome.;Beyond case studies of natural evolution, this dissertation also describes my efforts to better understand methods of ancestral sequence reconstruction. Such work consisted of computational analysis of an experimentally-derived data set in order to benchmark these methods as well as conducting simulations. In total, this particular computational work provides us with greater insights to the accuracies and limitations of ancestral sequence reconstruction methods.;The work presented in this dissertation highlights the diverse questions that evolutionary studies attempt to address and the different biological levels that can be studied to answer these questions.