Probing the aggregated effects of purifying selection per individual on 1,380 medical phenotypes in the UK Biobank

Understanding the relationship between natural selection and phenotypic variation has been a long-standing challenge in human population genetics. With emergence of biobank-scale datasets, along with new statistical metrics to approximate strength purifying at variant level, it is now possible correlate proxy individual relative fitness range medical phenotypes. We calculated per-individual deleterious load score by summing total number derived alleles per after incorporating weight that approximates selection. assessed four methods for weight, including GERP, phyloP, CADD, fitcons. By quantitatively tracking each these scores site frequency spectrum, we identified phyloP as most appropriate weight. The phyloP-weighted was then across 15,129,142 variants 335,161 individuals from UK Biobank tested association on 1,380 After accounting multiple test correction, observed strong amongst coding sites only 27 traits body mass, adiposity metabolic rate. further signals were driven common (derived allele > 5%) high (phyloP 2). Finally, through permutation analyses, showed had an excess nominally significant associations many These results suggest broad impact phenotypes highlight tool disentangle complex