Genetic variation in African populations: A multi-locus approach to understanding selection and demography in humans

Mutation, recombination, selection, and demographic processes (such as gene flow and genetic drift) have shaped genetic variation, but the relative impact of these evolutionary forces remains poorly understood. This problem motivates this study which examines three regions of the genome---beta-globin, the Y-chromosome, and mtDNA---in a two part approach to assess the relative impact of evolutionary forces on human genetic variation in Africa. The first approach characterizes levels of nucleotide variability and linkage disequilibrium across the beta-globin gene and recombinational hotspot in a sample of malarial-resistance alleles (HbC and HbS). Results suggest that the age of the HbC allele is <5,000 years and selection coefficients are 0.04-0.09 and, recombination is observed within 1-kb of the selected site on >1/3 of the chromosomes sampled ( Appendix A). A long-standing question regarding the HbS allele is whether it originated multiple times via recurrent mutation or whether it arose once and was transferred to different haplotypic backgrounds through recombination. These results indicate that recombination played a critical role in generating haplotypic diversity at beta-globin and can explain the origins of the Bantu and Senegalese HbS haplotypes (Appendix B). The second approach examines Y-chromosome and mtDNA variation to disentangle the relative effects of demographic forces. A detailed characterization of the Y-chromosome and mtDNA in >1000 individuals from ∼40 populations reveals that patterns of variation from these paternally- and maternally-inherited loci are remarkably different, suggesting that sex-specific demographic processes have influenced African genetic variation, particularly among agriculturalists (Appendix C). Hunter-gatherer populations carry a suite of Y-chromosomes that differ from those of agricultural populations. The examination of Y-SNP and Y-STR variation in eight hunter-gatherer populations reveals the presence of a very old, >50-kya, derived lineage (B2b) shared among these populations, which is absent in agricultural populations, suggesting that hunter-gatherer populations share an ancient common ancestry (Appendix D). Finally, the Y-chromosome results are placed into a broader evolutionary context in a phylogeographic summary as it relates to archeological and linguistic variation in Africa (Appendix E). Together these results underscore the vastly different effects that various evolutionary forces have had on shaping human genetic variation in Africa.