Adaptations to different environments are likely to play an important role in variation in disease prevalence among ethnic groups. Thus, an accurate characterization of local adaptations in humans is of fundamental importance to understanding disease susceptibility, as well as other phenotypes. Currently, it is thought that many local adaptations result from the dispersal of anatomically modern humans from East Africa. If so, patterns of polymorphism from non-African individuals should show the signature of adaptations dating to 40- 100 Kya. To date, however, scans of polymorphism data from a limited number of populations have yielded conflicting results as to both the chronology and geography of local adaptations. To clarify these issues, we propose to: 1) Generate new polymorphism data from non-coding regions and use it together with existing data to infer a sensible demographic model for each of 15 populations. These models will provide a framework within which to reliably assess the evidence for positive selection and estimate its timing. 2) Characterize the timing and geographic distribution of adaptations for a known selected phenotype linked to Out of Africa expansions into new environments, namely skin pigmentation. 3) Characterize the timing and geographic distribution of adaptations for unknown phenotypes in genes reported to have been under selection in one of the HapMap populations. With the rapid growth of polymorphism studies, our work will provide an interpretive framework for large- scale analysis of the signature of selection based on polymorphism data. Moreover, it will yield important insights into the demographic and selective factors that shape disease susceptibility loci.

R01GM079558

2007-03-01

2012-02-28