With the advent of new sequencing technologies, we now have generated more data than ever before. These data have greatly improved our knowledge of human history, human adaptation to different environments and human disease. Genome-wide studies have highlighted many genes or genomic loci that may play a role in adaptive or disease related phenotypes of biological importance. Now that we have access to thousands of human genomes from a diverse set of populations around the globe, we can zoom in at the local scale (e.g. within a gene) and leverage information from multiple populations to understand the observed patterns of genetic variation, without the ascertainment bias associated with the older array-based technologies. In addition, thanks to advances in DNA extraction and library preparation, we now are beginning to have access to DNA sequence data from ancient human samples. We propose to leverage these collections of modern and ancient sequence data to address some key questions in the field, and we have identified opportunities for methods development and biological discovery. The common theme in the proposal is to understand the different modes of natural selection in human populations. We plan to develop methods to detect shared selective events across populations by means of extending current statistical summaries, and methods for detecting admixture-facilitated adaptation which we believe is likely a common mode of natural selection based on our earlier published work. We will apply these tools to new datasets to characterize the interplay of natural selection, archaic and modern admixture in populations in the Americas and make a comparative analysis of modern and ancient European samples to understand the changing profile of medically important risk alleles for disease. As a result our work will reveal evolutionary processes that have played an important role in human evolution and disease.
If the long-term goal of human genomic research is to make personalized medicine a reality, then we need to be able to characterize each individual's genetic makeup. As we gather more data from both ancient and modern humans, one emergent feature of human history is that population interactions were more frequent than previously thought. The goal of this research project is to develop statistical methods that leverage multiple populations from the present and the past to boost our ability to address questions in human evolutionary biology that may overturn long-held assumptions about beneficial and deleterious mutations.