Open Collections

Jones, Bradley Robert

University of British Columbia

Human immunodeficiency virus 1 (HIV) maintains an ongoing pandemic. Despite combined antiretroviral therapies (cART) that can suppress viral load, halt progression to acquired immunodeficiency syndrome, and reduce morbidity, HIV infection has no durable cure. This is due in part to the existence of HIV proviruses that are integrated in CD4+ T cells in a transcriptionally latent state that can persist in the presence of cART. Forming a complete understanding of the dynamics of this “persistent” or “latent” reservoir is important for developing a durable HIV cure. There are some aspects of the persistent reservoir dynamics that are poorly understood, for example the timing of integration and duration of persistence of latent proviruses. In this thesis, I explore and develop different methods based on phylogenetics to estimate integration dates of proviral sequences. I start by describing a simulation of within-host HIV sequence evolution that includes the HIV persistent reservoir. This simulation is used to assess the accuracy of different integration date estimation methods. Next, I describe a method to estimate integration dates using linear regression on a maximum likelihood phylogeny and its web application implementation. I performed the linear regression on four participant data sets to estimate their proviral integration date distributions. These data included sequences that were collected from cells from different CD4+ T cell subsets. Finally, I explore a Bayesian method that uses a package I created for the software BEAST2 to estimate proviral integration dates. I applied this Bayesian method to one simulated data set and two empirical data sets. Overall, I found varied distributions of HIV integration dates among individuals living with HIV. This suggests that HIV cure interventions may benefit from personalized approaches. The simulations found that the method using Bayesian analysis produced the most accurate results. However, this method produced unrealistic results on some empirical data sets unless care was taken in selecting priors. Developing accurate and adaptable methods to understand the HIV persistent reservoir will assist HIV cure research.

Text

2023-08-29

Attribution-ShareAlike 4.0 International

http://hdl.handle.net/2429/85663

Bioinformatics

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-sa/4.0/