A research thesis submitted for the degree of Doctor of Philosophy at the University of Oxford.

J Parker

Funded by: Natural Environment Research Council (UK) with support from Linacre College, Oxford.

Abstract: This thesis examines the evolutionary biology of the RNA viruses, a diverse group of pathogens that cause significant diseases. The focus of this work is the relationship between the processes driving the evolution of virus populations within individual hosts and at the epidemic level.

First, Chapter One reviews the basic biology of RNA viruses, the current state of knowledge in relevant topics of evolutionary virology, and the principles that underlie the most commonly used methods in this thesis.

In Chapter Two, I develop and test a novel framework to estimate the significance of phylogeny-trait association in viral phylogenies. The method incorporates phylogenetic uncertainty through the use of posterior sets of trees (PST) produced in Bayesian MCMC analyses.

In Chapter Three, I conduct a comprehensive analysis of the substitution rate of hepatitis C virus (HCV) in within- and between-host data sets using a relaxed molecular clock. I find that within-host substitution rates are more rapid than previously appreciated, that heterotachy is rife in within-host data sets, and that selection is likely to be a primary driver.

In Chapter Four I apply the techniques developed in Chapter Two to successfully detect compartmentalization between peripheral blood and cervical tissues in a large data set of human immunodeficiency virus (HIV) patients. I propose that compartmentalization in the cervix is maintained by selection.

I extend the framework developed in Chapter Two in Chapter Five and explore the Type II error of the statistics used.

In Chapter Six I review the findings of this thesis and conclude with a general discussion of the relationship between within- and among-host evolution in viruses, and some of the limitations of current techniques.