Date Approved

7-20-2026

Embargo Period

7-20-2026

Document Type

Thesis

Degree Name

M.S. Pharmaceutical Sciences

Department

Pharmaceutical Sciences

College

College of Science & Mathematics

Advisor

Chun Wu, Ph.D.

Committee Member 1

Zhihong Wang, Ph.D.

Committee Member 2

Zwiwei Liu, Ph.D.

Abstract

Human immunodeficiency virus type 1 (HIV-1) is a single-stranded RNA retrovirus whose error-prone reverse transcriptase drives a high mutation rate, enabling adaptive mutations on drug target proteins that confer antiviral resistance and complicate long-term treatment. This study therefore aims to detect such adaptive mutations on HIV-1 drug target proteins using our novel analysis framework (c/µ) on 5,711 genomic sequences, with findings expected to inform timely therapeutic adjustments and guide the development of more robust antiretroviral strategies. In Chapter 1, our novel mathematical framework (c/µ) and the Near Neutral Unbalanced Selection Theory (NNUST), which posits that most mutations experience weak but non-neutral selection pressures, are introduced and applied to HIV-1. Our analysis of the molecular evolution of HIV-1 revealed that the (c/µ) of each nucleotide/amino acid site can be calculated to infer the overall site-specific selection fitness of the genome. In Chapter 2, the mutation rate µ is approximated independently of genetic codon tables and Markovian models by applying two complementary approaches: a time-based method using temporal sequence data and a position-based method leveraging genome-wide substitution rates. Using this framework, the c/µ test revealed an unexpected L-shaped distribution of fitness effects (DFE) and evidence of molecular clock violation across several genomic segments over 92 years of viral evolution. In Chapter 3, application of the codon-based c/µ method identified literature-validated adaptive mutations in Reverse Transcriptase, Protease, and Integrase, demonstrating the framework's ability to detect functionally relevant drug resistance mutations. Collectively, this work establishes a quantitative, multidisciplinary framework for understanding HIV-1 molecular evolution and provides insights that may inform future antiviral strategies and drug resistance detection efforts, as further validated through case studies on SARS-CoV-2 and EV68, demonstrating the framework's broader applicability to RNA viruses.

Download

Share

COinS