The role of admixture in human evolution

PROJECT SUMMARY / ABSTRACT The extent and relevance of admixture between human populations has remained one of the most hotly debated questions in human genomics. Recent studies have revealed that admixture between diverged populations has been a recurring feature of human evolution. However, despite the ubiquity of admixture in humans, it is typically not accounted for in population genomic studies of natural selection. This project will utilize ancient and modern human genetic variation data as well as reconstructed ancient microbiomes and pathogens to learn about the evolutionary consequences of mixing between human populations. First, this project will develop a novel statistical method that can detect genomic footprints of genetic adaptation (selective sweeps) in admixed ancient DNA, as well as estimate the selection coefficient and beneficial allele trajectory. This will allow us to accurately quantify and utilize adaptation signatures across different temporal samples and investigate the confounding effect of past admixture on selection signatures in modern samples. Further, we will contrast a model of homogeneous selection with a model of intense selection during the Out- of-Africa (OoA) migration event to explore if adaptation to colder European environments might have been a prerequisite to the OoA expansion. Second, we will evaluate the robustness of methods for detecting archaic adaptive introgression in modern human data, by examining an alternative hypothesis of passive genetic hitchhiking of archaic segments with mutations that originated in modern humans. This will allow a more accurate interpretation of introgression signatures in modern genetic variation data. Third, we will sequence archaeological skeletal samples from the ancient Levant that are bracketing a major historical event, the Late Bronze Age Collapse, which is characterized by extensive migration and cultural turnover. This study will provide important insights into the extent to which such a major admixture event was associated with changes in population size, health-related parameters of the microbiome, and the appearance of novel pathogens. Finally, we will develop a novel selection inference approach for modern admixed populations based on the Ancestral Recombination Graph (ARG). Using the ARG for population genetic inference promises to boost power compared to previous approaches and will allow us to accurately estimate the onset of selection. Knowledge of the onset of selection aids our understanding of possible selective pressures in cases where no ancient DNA of an ancestral pre-admixture population is available. In conclusion, the insights from this grant will be critical to our understanding of the evolutionary origin of human adaptations and the health implications of historical population admixture. The novel methods that we will develop will lead to an unprecedented level of resolution regarding the origin of genomic segments and mutations and when and why they had been selected. These insights will increase our ability to interpret individual genomic introgression and admixture profiles and thus contribute to the advancement of personalized genomic medicine.