Ancestry-Dependent Enrichment of Deleterious Homozygotes in Runs of Homozygosity

Abstract

Runs of homozygosity (ROH) are important genomic features that manifest when an individual inherits two haplotypes that are identical by descent. Their length distributions are informative about population history, and their genomic locations are useful for mapping recessive loci contributing to both Mendelian and complex disease risk. We have previously shown that ROH, and especially long ROH that are likely the result of recent parental relatedness, are enriched for homozygous deleterious coding variation in a worldwide sample of outbred individuals. However, the distribution of ROH in admixed populations and their relationship to deleterious homozygous genotypes is understudied. Here we analyze whole-genome sequencing data from 1,441 unrelated individuals from self-identified African American, Puerto Rican, and Mexican American populations. These populations are three-way admixed between European, African, and Native American ancestries and provide an opportunity to study the distribution of deleterious alleles partitioned by local ancestry and ROH. We re-capitulate previous findings that long ROH are enriched for deleterious variation genome-wide. We then partition by local ancestry and show that deleterious homozygotes arise at a higher rate when ROH overlap African ancestry segments than when they overlap European or Native American ancestry segments of the genome. These results suggest that, while ROH on any haplotype background are associated with an inflation of deleterious homozygous variation, African haplotype backgrounds may play a particularly important role in the genetic architecture of complex diseases for admixed individuals, highlighting the need for further study of these populations.

Strongly deleterious mutations are a primary determinant of extinction risk due to inbreeding depression

“…Although the insight that strongly deleterious mutations that are fully recessive can accumulate at greater frequencies in larger populations has been noted elsewhere (Nei 1968;Hedrick 2002;Hedrick and Garcia-Dorado 2016;Szpiech et al 2019), our results add to this work by demonstrating that that these effects persist under realistic genomic parameters and models of dominance, including both an hs relationship and our hmix model (Figure 4). This result is due to the fact that strongly deleterious mutations under these models tend to be highly recessive (Simmons and Crow 1977;Agrawal and Whitlock 2011;Huber et al 2018), and are therefore hidden from purifying selection when present as heterozygotes in large populations Moreover, when examining models with a single dominance coefficient for all mutations, we find that our results with fully recessive mutations (h=0 or 0.01) are highly similar to those under more realistic models of dominance, whereas the impact of ancestral demography is greatly diminished with partially recessive (h=0.05 or 0.2) mutations and nonexistent with additive (h=0.5) mutations (Figure 4).…”

SELAdb: A database of exonic variants in a Brazilian population referred to a quaternary medical center in São Paulo

“…By further quantifying this overlap using a neural network classifier and the subdivided AMR groups ( Figure 4B ), we could observe that the majority of individuals in SELAdb are classified as PUR (60.2%). This observation suggests a similar contribution of European and African ancestries to both PUR and the southeastern Brazilian populations, which is consistent with common historical aspects of colonization of both geographical regions and recent studies defining the ancestral contributions to each population ( 3 , 24 ).…”

The extent and impact of variation in ADME genes in sub-Saharan African populations

“…With high coverage data, we are able to detect ROHs of at least 300kb in size. High ROH is a measure of inbreeding decreased fitness and may be associated with ill health [26, 27]. However, ROH are not randomly distributed across the genome and islands of homozygosity (ROHi) are known to exist: regions where the ROH of several individuals within a population overlap [28].…”