PROJECT SUMMARY The long-term mission of my lab is to accurately model the effects of DNA sequence differences between indi- viduals on phenotypes relevant to disease and human evolution. Genetic variants that modify gene regulation are major contributors to both evolutionary divergence between species and differences in risk for disease among humans. However, due to the complexity of gene regulatory programs and the evolutionary histories that shaped patterns of genetic variation, interpreting the implications of an individual’s genome sequence re- mains challenging. This is a major roadblock to understanding the functional evolution of human-specific biolo- gy and mapping the genetic causes of disease. We integrate machine learning, large-scale functional genomics data, and ancient DNA to address this chal- lenge. The previous R35 funding enabled us with the flexibility to seize opportunities to make significant dis- coveries including: DNA sequence-based machine learning methods for predicting gene regulatory activity within and between species; demonstrating that selection against introgressed Neanderthal variation was the dominant trend across >400 traits, discovering the Neanderthal introgression introduced functional alleles lost in the out-of-Africa bottleneck; and developing a model for the functional evolution of gene regulatory elements. The previous R35 was the core source of support for work that led to 32 manuscripts, 26 of which my group led or co-led. Given our success in the previous grant period, my lab is uniquely well positioned with this renewal to build on our previous work to address fundamental questions in the following areas: 1. Developing powerful new machine learning methods that enable accurate prediction of individual-level molecular phenotypes, like gene expression, from DNA sequence. 2. Reconstructing molecular phenotypes of ancient humans and archaic hominins from their genomes to test hypotheses evolutionary transitions. 3. Leveraging new experimental technologies to dissect the functional drivers of cis vs. trans gene regula- tory divergence between species. 4. Interpreting non-protein-coding mutations identified in patient genomes to inform treatment and preven- tative care. Our work will produce much-needed methods for understanding the effects of genetic variant in gene regulato- ry regions and identifying mutations responsible for differences in disease risk between human populations. The renewal of our R35 will enable us to continue making significant contributions to these essential basic sci- ence and clinical challenges.