Viruses are the most abundant biological entities on Earth, infecting all types of cellular organisms. Yet less than 1% of the virosphere on our planet has been characterized to date. Viruses are both an important driver of bacterial evolution and have significant implications for human health, therefore understanding the relative contributions of various evolutionary forces in shaping their genomic landscapes is of critical importance both mechanistically as well as clinically. In my thesis I use computational genomic approaches to gain novel insights into bacteriophage and human cytomegalovirus evolution. In my first two chapters and associated appendices I characterized the complete genomes of the Cluster P bacteriophage Phegasus and Cluster DR bacteriophage BiggityBass, whose isolation hosts were Mycobacterium smegmatis mc²155 and Gordonia terrae CAG3, respectively. I also determined the bacteriophages' phylogenetic placement and computationally inferred their putative host ranges. For my fourth chapter I assessed the performance of several of these computational host range prediction tools using a dataset of bacteriophages whose host ranges have been experimentally validated. Finally, in my fifth chapter I reviewed the key parameters for developing an evolutionary baseline model of another virus, human cytomegalovirus.

Studying human genetic variation opens the possibility of understanding the details of population migrations, how humans develop and function, and why they get sick. To fully understand these things, genetic variation must be comprehensively characterized across globally diverse human populations and evolutionary knowledge can be used to inform studies of disease. In my dissertation I use computational methods to study human genetic variation. Each of my dissertation chapters focuses on a unique topic in the field of human evolutionary genetics. In the first chapter, I present PopInf, a computational pipeline to visualize principal components analysis output and assign ancestry to samples with unknown genetic ancestry, given a reference population panel of known origins. This pipeline facilitates visualization and identification of genetic ancestry across samples, so that this ancestry can be accounted for in studies of health and disease risk. In the next chapter, I investigate factors that shape patterns of genetic variation within and among four small-scale pastoral populations in northern Kenya. I find that geography predominantly shapes patterns of genetic variation in northern Kenyan human populations. In the next chapter, I investigate the extent to which Neanderthal introgression impacts liver cancer etiology. I find a pattern of overall enrichment of somatic mutations on Neanderthal introgressed haplotypes. Finally, through simulations, I investigate the effects of standard autosomal versus sex chromosome complement-informed alignment, variant calling and variant filtering strategies on variants called on the human sex chromosomes. I show that aligning to a reference genome informed on the sex chromosome complement of samples improves variant calling on the sex chromosome compared to aligning to a default reference, and variant calling is improved in males when calling the sex chromosomes haploid rather than diploid and when using haploid-based thresholds for filtering variants on the sex chromosomes. I provide recommendations for alignment, variant calling and filtering on the sex chromosomes based on these findings.