Study population and setting
This study explored the evolutionary origin of a stretch of DNA sequence, located on human chromosome 3, which has been consistently identified as associated with COVID-19 outcomes via large genome-wide association studies (GWAS). Using summary statistics from GWAS efforts of the Covid-19 Host Genetics Initiative and Severe Covid-19 GWAS Group, the single nucleotide polymorphisms (SNPs) in this chromosome 3 region most strongly associated with COVID-19 outcomes were compared across whole-genome sequences of extant human populations, Homo neanderthalis, and Homo denisova to determine the origin of this COVID-19 associated region.
Summary of Main Findings
The most significantly associated COVID-19 SNPs within the chromosome 3 locus were identified to be highly associated with one another, i.e., in strong linkage disequilibrium (LD) (r2>0.98), a haplotype within a ‘core’ region 49.4kb long as well as a wider 333.8kb haplotype region of weaker LD (r2>0.32). These haplotypes were determined to have arisen via gene flow from Neanderthals, with prevalence estimates made using the 1000 Genomes Project’s South Asian (30%), European (8%), admixed American (4%), East Asian (0%), and African (0%) populations. While the authors suggest the high prevalence of this haplotype among those of South Asian ancestry in the UK could account for an increased risk of death in this population and claim this haplotype is currently under purifying selection (i.e., being actively removed from the population) due to SARS-CoV-2, no evidence is presented to support these claims.
This study leverages high quality Neanderthal and Denisovan genomes to determine the evolutionary origin of a locus associated with COVID-19 disease. This study also utilizes well-described population-based allele and genotype frequency databases to support their findings.
This study leverages the most significantly associated SNPs identified by two GWAS of COVID-19 disease and is therefore constrained by the study designs and phenotype definitions used in these GWAS; as rs11385942 and rs35044562 were identified by comparing severe COVID-19 cases to population-based controls of unknown SARS-CoV-2 exposure outcomes, some control individuals could have developed severe COVID-19 had they been exposed potentially resulting in alternative index SNPs being identified in future analyses. Further, the increased risk of death due to COVID-19 among those of Bangladeshi ethnicity living in the UK compared to those of ‘white British ethnicity’, used to support the functional importance of this Neanderthal-derived haplotype, is misleading as it does not account for disparities in comorbidities which ameliorate these differences in risk of death. The authors also state the identified Neanderthal haplotype is absent from African populations, yet the index variant rs11385942 occurs in 2%-7% of the 1000 Genomes African sub-populations, a potential contradiction. Finally, the authors provide no evidence supporting their substantial claim that this haplotype is currently under purifying selection due to SARS-CoV-2.
The identification of a Neanderthal-derived haplotype as being the primary source of COVID-19 associated variants is intriguing, as Neanderthal-derived genetic loci can be enriched for genetic factors related to the immune response.
This review was posted on: 29 October 2020