Identifying signatures of positive selection in human populations from North Africa

Abstract

Because of its location, North Africa (NA) has witnessed continuous demographic movements with an impact on the genomes of present-day human populations. Genomic data describe a complex scenario with varying proportions of at least four main ancestry components: Maghrebi, Middle Eastern-, European-, and West-and-East-African-like. However, the footprint of positive selection in NA has not been studied. Here, we compile genome-wide genotyping data from 190 North Africans and individuals from surrounding populations, investigate for signatures of positive selection using allele frequencies and linkage disequilibrium-based methods and infer ancestry proportions to discern adaptive admixture from post-admixture selection events. Our results show private candidate genes for selection in NA involved in insulin processing (KIF5A), immune function (KIF5A, IL1RN, TLR3), and haemoglobin phenotypes (BCL11A). We also detect signatures of positive selection related to skin pigmentation (SLC24A5, KITLG), and immunity function (IL1R1, CD44, JAK1) shared with European populations and candidate genes associated with haemoglobin phenotypes (HPSE2, HBE1, HBG2), other immune-related (DOCK2) traits, and insulin processing (GLIS3) traits shared with West and East African populations. Finally, the SLC8A1 gene, which codifies for a sodium-calcium exchanger, was the only candidate identified under post-admixture selection in Western NA.

Figure 1

Population structure analysis of the North African dataset. (a) PCA of North African individuals (full coloured symbols) and reference populations (empty symbols) from West and East Africa (YRI and LWK, respectively), Europe (CEU), and the Middle East. (b) ADMIXTURE analysis at K = 4 with the same samples. YRI, Yoruba from Ibadan (Nigeria); LWK, Luhya in Webuye (Kenya); CEU, Utah residents with North- and West- European ancestry.

Figure 2

Local Ancestry Deviation (LAD) identified in chromosome 2 of Western North African populations (NAW). The horizontal lines mark the threshold of |LAD| above 4.42, indicative of putative post-admixture selection. A zoom in of the genes contained within the genomic region surpassing the defined threshold is also represented as extracted from the UCSC Genome Browser on Human (GRCh37/hg19).

Figure 3

Signals of positive selection identified in North African (NA) populations that are shared with European or West and East African (WEA) populations. (a) |XP-EHH| scores for the NA and European populations (when compared against WEA) in chr15:48,261,821–48,561,821, which contains the SLC24A5 gene. (b) |XP-EHH| scores for the NA and European populations (when compared against WEA) in chr12:88,736,570–89,036,570, which comprises the KITLG gene. (c) |iHS| scores for Western NA and WEA in chr11:5,361,220–5,661,220, where the HBE1 and HBG2 genes reside. The horizontal dashed lines represent the minimum threshold of significance for each statistic. Within each genomic region, gene tracks were extracted from the UCSC Genome Browser on Human (GRCh37/hg19).