Project Summary/Abstract Obstructive sleep apnea (OSA) is a common breathing disorder characterized by recurrent episodes of upper airway collapse with reduced airflow, oxygen desaturation and consequent intermittent hypoxia, which are implicated in increased risk of cardiovascular disease. Surprisingly, although there is a strong dose-response association between OSA and stroke, evidence of a link between OSA and coronary artery disease (CAD) is less certain. One possibility is that OSA is a “double-edged sword” with some positive as well as negative consequences for CAD event risk. At certain doses, intermittent hypoxia (IH) may reduce CAD event risk by ischemic preconditioning, with positive effects on angiogenesis, coronary collateralization and increased antioxidant production. By contrast, IH may increase CAD event risk via oxidative stress and inflammation. Consequently, we investigate the following scientific questions: 1) which specific features of IH give rise to CAD event risk, and 2) how does IH influence or depend on prior genetic risk to impact CAD. Therefore, we aim first to investigate dose-dependent effects of IH on CAD, and second to account for IH as an effect modifier of genetic risk of CAD, which may vary by genetic regulatory pathway in a dose-dependent fashion. This will require us to develop new methods to better quantify both IH severity and pathway-specific genetic risk of CAD events and to interrogate interactions of these two sources of risk, thereby clarifying the relationship between OSA and CAD. Specifically, we hypothesize that among susceptible individuals with OSA, IH will further perturb pathways already under genetic dysregulation and lead to increased CAD event risk. Conversely, in some cases exposure to OSA-related IH will reduce CAD event risk, for example by overdriving angiogenic or antioxidant pathways. To investigate pathway-level genetic interactions, we capitalize on recent advances in polygenic risk scores (PGRS) and state-of-the-art functional genetics to assign single nucleotide variant (SNV) effects to genes and genes to pathways, computing novel pathway- specific polygenic risk score (PS-PGRS) for each subject within each pathway of interest. Hence our work addresses questions of fundamental scientific interest, clarifying pathways through which IH affects CAD, as well as how risks interact to predict downstream CAD consequences in individual patients. Investigating these scientific questions and following the training program proposed in this fellowship application will allow me to enhance my biostatistics training in the area of risk and severity score methods development while learning to work with complex applied problems in deeply-phenotyped large-scale multi-ancestry cohorts. In turn this will advance me toward my career goal of becoming an independent researcher focusing on incorporating biological and clinically relevant data into statistical models needed to understand disease heterogenei...