Fig. 4: Geographic and ancestry-related contributions to genetic variation in the MXB.
From: Clinical genetic variation across Hispanic populations in the Mexican Biobank
a, UpSet plot summarizing results from a GLM assessing the association of genetic ancestry (Indigenous, African) and geographic variables (latitude, longitude) with the distribution of 42,769 clinically relevant variants. Vertical bars represent the number of SNPs significantly associated with each term or combination of terms (P < 0.05); horizontal bars indicate the total number of variants associated with each variable. Combinations accounting for <1% of variants are not labeled. b–e, Geographic distribution of allele frequencies for four representative SNPs in the MXB, each illustrating a distinct pattern of genetic variation. Star indicates that the Indigenous ancestry component can be interpreted as a mirror of European ancestry in the opposite direction, due to their negative correlation. b, rs2242480 (fentanyl metabolism), enriched in individuals with higher Indigenous ancestry. c, rs45613035 (asthma risk), showing a longitudinal gradient with elevated frequencies in eastern Mexico. d, rs2239785 (apolipoprotein L1, kidney disease), associated with African ancestry. e, rs7147503 (body mass index), exhibiting a latitudinal gradient with increased frequency in the south. Images in b–e are reproduced with permission from Valle-Jones D. (ref. 38), used under a free use license. f, Boxplots of the absolute effect size from the GLM for each variable, stratified by direction (positive versus negative effect). The Indigenous ancestry term shows significantly stronger negative effects compared to positive ones (t test two-sided, ****P ≤ 0.001; NS, not significant, P > 0.05. Indigenous ancestry, t(33,887) = 79.1, P < 1 × 10−300; longitude, t(9,316) = 0.99, P = 0.323; African ancestry, t(6,888) = –36.9, P = 6.9 × 10−272; latitude, t(2,588) = 1.92, P = 0.055), suggesting a systematic bias in allele frequency patterns toward European enrichment.
