Project Summary/Abstract The use of combination antiretroviral therapy (cART) has significantly reduced HIV-related morbidity and mortality. However, cART may exacerbate the central nervous system (CNS)-related adverse effects on mental health for people with HIV (PWH). These adverse effects may result in ART discontinuation with undesirable downstream consequences such as HIV disease progression, decreased health outcomes, and increased likelihood of HIV transmission, causing public health concerns. Depression is the most frequently reported mental health comorbidity caused by CNS injury in PWH, with prevalence ranging from 20% to 60%. Understanding factors (e.g., drug-drug interactions) contributing to ART-related depression is critical and remains a high priority NIMH research area. In addition, since PWH must continue cART indefinitely, optimizing sequential cART treatments over a long-time span tailored to individuals’ evolving clinical characteristics and treatment histories is important for improving long-term mental health for PWH. However, there are numerous possible drug combinations with complicated drug-drug interactions and thus creating complex data patterns, such as heterogeneity, high-dimensionality, and sparseness, making it highly challenging to develop appropriate statistical models for these problems - which is a critical gap we aim to fill. This proposal will leverage large public HIV datasets, including Women's Interagency HIV Study (WIHS), to develop data-driven approaches to facilitate deciphering cART-depression relations and guide more effective cART treatments. This proposal is organized into three aims: 1) Develop Bayesian methods to learn longitudinal cART effects on depression and investigate effect modifiers (e.g., polymorphic drug metabolism, aging); 2) Develop Bayesian decision frameworks to optimize personalized sequential cART assignments with the goal of improving long-term mental health outcomes for PWH; (3) encapsulate statistical methods and computational algorithms into user-friendly open-source software for practical use, clinical translation, and dissemination. Findings from this study are expected to expand our understanding of cART effects on depression, and have potential clinical utility to facilitate precision medicine in HIV.