Current approaches to precision medicine focus on a patient’s clinical history and are often combined with known genetic risk factors. Multiple diseases have shown strong prevalence differences across racial groups and as a result race has been incorporated into clinical practice as a factor to consider in personalizing a treatment plan. However, multiple studies have shown that administratively determined race or self-reported race are imprecise estimates of an individual’s actual genetic ancestry. Imprecise racial/ancestral identification may contribute to lack of positive response to a personalized treatment plan. Furthermore, recent work by several groups has shown that for some diseases genetic ancestry (i.e., global ancestry) may directly interact with a patient’s clinical characteristics to modify disease risk and that this interaction varies at specific points in their genome (i.e., local ancestry). Within this proposal we leverage the rich phenotypic information available from large electronic health record (EHR) biobanks to comprehensively evaluate the relationship between disease risk, drug response, and genetic ancestry and identify specific clinical characteristics of patients that interact with global and local genetic ancestry. Additionally, we will use 1000 genomes to not only map disease risk to specific genetic ancestries but use this information to better understand geographic origins of disease.