Clusters of adaptive evolution in the human genome

Abstract

Considerable work has been devoted to identifying regions of the human genome that have been subjected to recent positive selection. Although detailed follow-up studies of putatively selected regions are critical for a deeper understanding of human evolutionary history, such studies have received comparably less attention. Recently, we have shown that ALMS1 has been the target of recent positive selection acting on standing variation in Eurasian populations. Here, we describe a careful follow-up analysis of genetic variation across the ALMS1 region, which unexpectedly revealed a cluster of substrates of positive selection. Specifically, through the analysis of SNP data from the HapMap and Human Genome Diversity Project-Centre d'Etude du Polymorphisme Humain samples as well sequence data from the region, we find compelling evidence for three independent and distinct signals of recent positive selection across this 3 Mb region surrounding ALMS1. Moreover, we analyzed the HapMap data to identify other putative clusters of independent selective events and conservatively discovered 19 additional clusters of adaptive evolution. This work has important implications for the interpretation of genome-scans for positive selection in humans and more broadly contributes to a better understanding of how recent positive selection has shaped genetic variation across the human genome.

Keywords: clustered adaptive events.

Figure 1

Patterns of population specific F_ST and heterozygosity at chromosome 2 (p13.3–p13.1). The location of each gene (going from left to right: SEC15L2, ALMS, GCS1) is marked by a gray rectangle and black arrow, and additional genes located in the region are shown as white rectangles. Previously inferred recombination hotspots are denoted by black rectangles. For each HapMap sample, population specific F_ST and heterozygosity are plotted individually in orange and blue respectively.

Figure 2

Distribution of SEC15L2 alleles in 52 populations. Haplogroup frequencies are indicated with pie charts. The ancestral allele is shown in white, and the derived allele is shown in black.

Figure 3

Distribution of ALMS1 alleles in 52 populations. Haplogroup frequencies are indicated with pie charts. The ancestral allele is shown in white, and the derived allele is shown in black.

Figure 4

Distribution of GCS1 alleles in 52 populations. Haplogroup frequencies are indicated with pie charts. The ancestral allele is shown in white, and the derived allele is shown in black.

Figure 5

Phylogenetic relationships among seven continental groups. Maximum likelihood trees for each of the three genes are shown rooted with chimpanzee as the outgroup. The length of the scale bar corresponds to 1% of sites differing between haplotypes.

Figure 6

Patterns of population specific F_ST at chromosome 2 (q12.3–q13). The location of EDAR is marked by a gray rectangle and black arrow, and additional genes located in the region are shown as white rectangles. Previously inferred recombination hotspots are denoted by black rectangles. For each HapMap sample, population specific F_ST is plotted individually in blue, and the dashed line in each panel indicates the 99th percentile for each population specific F_ST. Gray shading indicates 100 kb windows with significantly elevated average values of population specific F_ST (note, the apparent width of some shaded regions vary because adjacent windows are significant).