EVOLUTIONARY BENEFITS OF LATENCY IN WITHIN-HOST HIV INFECTION DYNAMICS
HIV is a retrovirus that infects helper T cells (CD4 + T cells) in the human immune system. At the cellular scale, HIV generates both actively and latently infected cells. Ac- tively infected cells produce mature virions and are often the primary target of antiretroviral therapies. In contrast, latently infected T cells can do not produce virus particles, are hard to detect and treat, and can be reactivated to produce new virions. Understanding the dy- namics of latent infections is critical to the development of strategies to treat and control the spread of HIV. In this thesis, we study a variant of within-host models of HIV infection dynamics in- cluding proliferation of both susceptible and latently infected CD4+ cells. In this model, HIV infection of susceptible cells can result in acute or latent infections. The key innova- tion here is to identify the relative contributions of the active and latent pathways towards viral fitness, both in the initial and later stages of the within-host dynamics. We do so by leveraging a new approach to decomposing viral fitness developed in the context of phage- bacteria interactions. Our work highlights how variation in susceptible cell densities, viral life history traits, and retroviral therapies jointly influence dynamic selection pressures for active and latent infections.