Quantifying the impact of rare and ultra-rare coding variation across the phenotypic spectrum

Abstract

Protein truncating variants (PTVs) are likely to modify gene function and have been linked to hundreds of Mendelian disorders. However, the impact of PTVs on complex traits has been limited by the available sample size of whole-exome sequencing studies (WES). Here we assemble WES data from 100,304 individuals to quantify the impact of rare PTVs on 13 quantitative traits and 10 diseases. We focus on those PTVs that occur in PTV-intolerant (PI) genes, as these are more likely to be pathogenic. Carriers of at least one PI-PTV were found to have an increased risk of autism, schizophrenia, bipolar disorder, intellectual disability and ADHD (P-value (p) range: 5x10-3-9x10-12). In controls, without these disorders, we found that this burden associated with increased risk of mental, behavioral and neurodevelopmental disorders as captured by electronic health record information. Furthermore, carriers of PI-PTVs tended to be shorter (p=2x10-5), have fewer years of education (p=2x10-4) and be younger (p=2x10-7); the latter observation possibly reflecting reduced survival or study participation. While other gene-sets derived from in vivo experiments did not show any associations with PTV-burden, gene sets implicated in GWAS of cardiovascular-related traits and inflammatory bowel disease showed a significant PTV-burden with corresponding traits, mainly driven by established genes involved in familial forms of these disorders. We leveraged population health registries from 14,117 individuals to study the phenome-wide impact of PI-PTVs and identified an increase in the number of hospital visits among PI-PTV carriers. In conclusion, we provide the most thorough investigation to date of the impact of rare deleterious coding variants on complex traits, suggesting widespread pleiotropic risk.