Systematic discovery of gene-environment interactions underlying the human plasma proteome in UK Biobank

Understanding how gene-environment interactions (GEIs) influence the circulating proteome could aid in biomarker discovery and validation. The presence of GEIs can be inferred from single nucleotide polymorphisms that associate with phenotypic variability - termed variance quantitative trait loci (vQTLs). Here, vQTL association studies are performed on plasma levels of 1,468 proteins in 53,752 UK Biobank participants. A set of 683 independent vQTLs are identified across 571 proteins, all of which are newly discovered. They include 65 variants that lack conventional additive main effects on protein levels. Over 1,400 GEIs are identified between 142 proteins and 101 lifestyle and metabolic exposures. GEI analyses uncover biological mechanisms that explain why 13/65 vQTL-only sites lack corresponding main effects. Stratified analyses also highlight how age, sex and genotype can interact to modify relationships between biomarkers and health-related traits. This study establishes the most comprehensive database yet of vQTLs and GEIs for the human proteome.