Genetic landscape of an in vivo protein interactome

SUMMARY Protein-protein interaction (PPI) networks accurately map environmental perturbations to their molecular consequences in cells, but effects of genome-wide genetic variation on PPIs remain unknown. We hypothesized that PPI networks integrate genetic and environmental effects, potentially defining biochemical mechanisms underlying complex polygenic traits. Here, we measured 61 PPIs in inbred strains of Saccharomyces cerevisiae with ∼12,000 single-nucleotide polymorphisms (SNPs) across the genome. Unlike mRNA expression and protein abundance that are primarily affected by SNPs local (in “cis”) to a gene, PPIs are predominantly affected by SNPs far (in “trans”) to the genomic loci of the interacting proteins. However, consistent with the PPI network’s small-world characteristic, these transacting SNPs are in neighboring genes in the network. We likewise discovered SNPs in non-coding RNAs and post-transcriptional regulators (3’ UTRs) with, counterintuitively, larger PPI-modulating effects than SNPs within protein-coding regions. Finally, we inferred known and novel mechanisms of action for yeast and human drugs. HIGHLIGHTS Protein-interaction quantitative trait locus (“piQTL”) mapping reveals sensitivity of in vivo PPIs to polymorphisms across the yeast genome Trans-piQTLs significantly outnumber and are stronger than cis-piQTLs SNPs in non-coding RNAs and 3’ UTRs have comparable effects to PPI as SNPs in coding regions piQTL mapping reveals known and novel mechanism of yeast and human drugs