Strong Selection at MHC in Mexicans since Admixture

Abstract

Mexicans are a recent admixture of Amerindians, Europeans, and Africans. We performed local ancestry analysis of Mexican samples from two genome-wide association studies obtained from dbGaP, and discovered that at the MHC region Mexicans have excessive African ancestral alleles compared to the rest of the genome, which is the hallmark of recent selection for admixed samples. The estimated selection coefficients are 0.05 and 0.07 for two datasets, which put our finding among the strongest known selections observed in humans, namely, lactase selection in northern Europeans and sickle-cell trait in Africans. Using inaccurate Amerindian training samples was a major concern for the credibility of previously reported selection signals in Latinos. Taking advantage of the flexibility of our statistical model, we devised a model fitting technique that can learn Amerindian ancestral haplotype from the admixed samples, which allows us to infer local ancestries for Mexicans using only European and African training samples. The strong selection signal at the MHC remains without Amerindian training samples. Finally, we note that medical history studies suggest such a strong selection at MHC is plausible in Mexicans.

Fig 1. Global ancestry proportions and principal components.

(a) and (b) are triangular plots for Viva and Lipid respectively. To produce a triangular plot, note that each individual associates a triplet of ancestry proportions (x, y, z) such that x + y + z = 1, and a unique point can be determined such that within an equilateral triangle its distances to three edges are x, y and z. (c) and (d) are PC plots for Viva and Lipid respectively. The PC plots shown are mirror images of the original as indicated by “–” sign in labels.

Fig 2. African average dosages.

Plot shows all 22 autosomes for two GWAS datasets. The spike at MHC region on chromosome 6 is rather striking in both datasets. The blue lines are the genome-wide mean of average dosages; the gray lines are mean ± 4ssd (ssd stands for sample standard deviation).

Fig 3. Comparison between different European and African training samples.

The comparison was performed with chromosome 6 of Lipid dataset. African (a) and European (b) average dosages for five sets of training samples shown in legend, where ALL means CEU+TSI−YRI+MKK−MAYA. (c) The difference of estimated European average dosages of Mexicans between two European training samples (see main text for explanation). (d) The violin plots of structure analysis of five HapMap3 populations, where ASW denotes Americans from the Southwest, an African American population. On each violin plot, gray dot denotes the median and black dot the mean.

Fig 4. Comparison of estimations with and without Amerindian training samples.

(a) African average dosages of Viva. (b) Amerindian average dosages of Viva. (c) African average dosages of Lipid. (d) Amerindian dosages of Lipid. We combined CEU and TSI as European training samples, and YRI and MKK as African training samples.