Studying Within-Host Viral Evolution Using Pooled Next-Generation Sequencing Data

Pooled next-generation sequencing allows multiple genomes to be sequenced at once in a single sample, with the resultant single nucleotide polymorphism data giving reliable estimates of allele frequencies and population genetic parameters in a cost-effective manner. This approach has potentiated new opportunities for understanding the evolution of virus populations within individual hosts over the course of infection, where the sequencing of individual genomes is exceedingly difficult and impractical. However, evolutionary tools for analyzing the latest forms of pooled-sequencing data have been lacking. In this thesis, I first review next-generation sequencing and relevant molecular evolution topics, including the unique features of RNA viruses. I conclude that viruses, given their extremely fast replication rates and within-host population sizes, are ideal models for studying evolution by natural selection. Next, simple methods are devised for estimating nonsynonymous and synonymous nucleotide diversity from pooled nextgeneration sequencing data, without the need for inferring linkage. I introduce SNPGenie, a new bioinformatics tool for applying these methods to any pooled or individual variant data. Finally, I use SNPGenie to address topics of both practical and theoretical interest in the evolution of simian hemorrhagic fever viruses (Arteriviridae) infecting red colobus monkeys (Procolobus rufomitratus tephrosceles), including fundamental questions regarding the effective population sizes of, the mutation rates experienced by, and the modes and efficacy of natural selection acting on within-host viral populations