PROJECT SUMMARY/ABSTRACT Although individually rare, Mendelian diseases are collectively common. Nearly 1% of people have a medical condition that can be traced back to a single gene. This is particularly true for patients with cardiovascular disease. Genetic testing is now commonly employed in clinical practice. As a result, it has become clear that we have an incomplete understanding of how genetic mutations cause Mendelian disease. Some patients with deleterious mutations display very severe symptoms while others are almost entirely unaffected. This is true even for childhood-onset disorders with severe cardiovascular symptoms, like Marfan Syndrome. Understanding the range of disease severity has important implications for diagnosis, prognosis, and management. In the past, cohort studies and case series have been used to gather this type of information, but these yield incomplete and biased views of disease heterogeneity. Therefore, new methods for studying Mendelian disease genetic and phenotypic diversity are urgently needed. Population-scale biobanks linked to electronic health records (EHRs) can provide a less biased view of genotype-to-phenotype relationships. Subjects are included in these datasets regardless of their medical history. EHR data also provides detailed phenotypic information on each subject. Finally, these biobanks now include hundreds of thousands of individuals, capturing rare genetic variation on an unprecedented scale. As a result, we hypothesize that population-scale biobanks can provide new insight into the genotype-to-phenotype relationships that underlie congenital cardiovascular syndromes (CCSs). This hypothesis will be tested in two specific aims. In Aim 1, we will use biobanks to develop quantitative scores that reproducibly summarize CCS- related phenotypic severity. These traits have multiple applications. In Aim 2, we will use them to build computational models that predict the phenotypic effects of CCS-related rare variants directly from sequence context. Once validated, these models should reduce diagnostic uncertainty in clinical practice. I am a clinical geneticist and physician-scientist devoted to improving the quality of healthcare provided to Mendelian disease patients. Long term, I plan to develop an independent research program that uses complex clinical and genetic datasets to improve our understanding of Mendelian disease risk, variability, and progression. My K38 research proposal is entirely consistent with these goals. In addition, it will provide valuable career development. New technical skills in EHR data analysis and statistical genetics will be acquired, as will academic skills like grant writing. During the award, I will be mentored by leaders in the fields of biomedical data science and human genetics, including Dr. Atul Butte and Dr. Neil Risch. Finally, the K38 award will serve as springboard for future funding opportunities and research independence. Therefore, the K38 StARRTS award...