Approved research

Estimates of allele age and impact in large data sets

Lay summary

The research will apply a new method for estimating how long ago a mutation has occurred in UK BIOBANK exome and genotype data. Knowledge of the age of a mutation informs on how harmful a mutation is, because a mutation that is younger than expected, given its frequency, will probably be harmful in its effects on health and fitness. This is particularly true for rarer mutations. It is for this reason that the UK BIOBANK offers an exceptional resource in its large exome and genotype data set. We will estimate the age of all the rare mutations in the data set. Because each gene will have many rare alleles, and because we can assess the overall impact of mutations on each gene (by our estimates of allele age), we can develop a detailed portrait of mutational harm and evolutionary constraint for each gene. We will contrast these results with estimates of phylogenetic constraint, which may reveal recent changes in how evolution has acted on individual genes. We will then join our estimates of allele age with results from mapping studies that quantify contributions of alleles and genes to particular phenotypes. By bringing these two types of analyses together, we can assess the correspondence between phenotypic effect and very recent evolutionary constraint. We can compare the selective burden (as estimated by mutation age) for genes that are known to carry common mutations (from GWAS studies) that contribute to negative health outcomes. These analyses will provide an alternative kind of insight for elucidating the genetic basis of variation in human health. Mapping studies have had to focus on relatively common alleles in order to establish statistical associations with particular phenotypes. Our analysis will complement these studies, by assessing the overall selective impact of the very rarest alleles. While we will not be assessing the impact of mutations on specific phenotypes, we can join our results with those from other studies that do focus on phenotypes.