Genome sequencing has seen broad interest in its potential to identify disease risk, but the strength of connections between genotypes and phenotypes is unclear.
In a study with results presented in JAMA Dermatology, researchers examined an autosomal recessive genetic condition, xeroderma pigmentosum (XP), as a model for evaluating genotype-phenotype relationships.
XP can result from mutations in certain DNA-repair genes. Between January 1, 2017, and May 4, 2018, the researchers reviewed relevant genetic data from patients with XP from National Institutes of Health (NIH) and Human Gene Mutation Database (HGMD) sources. They compared these data with alleles from 3 large genomic databases to estimate the rate of XP to be expected in the broader population.
The researchers identified 156 nonsense and missense mutations in 8 XP-associated genes from NIH and HGMD data and found frequencies for 65 of these in the Genome Aggregation Database (gnomAD). Two relatively common mutant alleles, XPF (ERCC4) p.P379S and XPC p.P334H, showed frequencies of 0.41% and 0.30%, respectively, in gnomAD.
The researchers used the Hardy-Weinberg equation to estimate the prevalence of homozygosity for these 2 alleles that should be present in the broader population, based on gnomAD frequencies. They found that there should be in excess of 8000 patients in the United States showing homozygosity for either of these mutations, and, thus, clinical XP. However, only 3 patients with XP have been identified with XPF p.P379S and 1 with XPC p.P334H.
Because of the great mismatch between genetic variants and clinical manifestations in this study, the authors urged caution when attempting to use genomic databases to tie disease risk to genetic variation.
Pugh J, Khan SG, Tamura D, et al. Use of big data to estimate prevalence of defective DNA repair variants in the US population [published online December 5, 2018]. JAMA Dermatol. doi: 10.1001/jamadermatol.2018.4473