Abstract Atrial fibrillation (AF) is of major public health importance because of increasing prevalence, and high lifetime risk, costs, morbidity, and mortality. Current AF therapies have partial efficacy, moderate adherence, high cost, and substantial morbidity. Hence, there is a profound need to develop a more comprehensive understanding of the etiology of AF to identify individuals at risk for AF and novel drug targets for AF therapies. In 2008, we organized the AFGen Consortium (AFGen) and since that time our highly collaborative, international consortium has led the field of AF genetics. We have described the vast majority of the more than 130 genetic loci that have been associated with AF. To complement our genome wide association data, we also have led efforts to perform whole exome (WES) and genome sequencing (WGS) in individuals with AF. We have identified loss of function variants in the sarcomeric protein, titin, that are significantly associated with early-onset AF. In our competitive renewal application, we now seek to extend our prior work in 4 directions. In Aim 1, we propose to conduct one of the largest disease-based analyses of WES and WGS data. In aggregate we will include over 91K AF cases and 769K controls from the NHLBI Trans-Omics for Precision Medicine program, the NHGRI Center for Common Disease Genomics program, the UK Biobank, 5 clinical trials from the TIMI Study, and the All of Us Program. Our primary analysis will be focused on the identification of AF associated genes. We will then take advantage of this multi-ancestry data to a) fine map loci to identify causal variants, b) develop polygenic risk scores that identify individuals at high risk across ancestries, and c) use Mendelian Randomization to assess both causal effects of risk factors on AF, and AF on heart failure and stroke. In Aim 2, we propose to use deep learning models to reconstruct left atrial (LA) size and function in the United Kingdom Biobank cardiac MRI imaging data in 100K individuals. In preliminary studies, we have derived measurements for 7 LA traits, identified more than 20 genetic loci associated with LA traits, and co-localized LA and AF risk genes. In Aim 3, we propose to validate the top 50 AF and LA associated genes by performing gene perturbation assays in stem cell derived atrial cardiomyocytes. We will perform gene knockouts or over expression followed by assays of myocyte structure and electrophysiology using high content imaging. Finally, in Aim 4, we will support the ongoing efforts of the AFGen Consortium and continuing training of early-stage investigators in a virtual fellowship. We believe that our multidisciplinary team brings together extraordinary expertise in AF genetics, epidemiology, bioinformatics, biostatistics, and basic cardiovascular research. Ultimately, we anticipate that our work will provide novel targets for the risk stratification, prevention, and treatment of AF.