Presented by

Dr. Karsten Suhre

Genome-wide association studies with proteomics are increasingly used to identify protein quantitative traits (pQTLs) that overlap with genetic loci of clinical interest and can inform decision making in drug target prioritization and drug repurposing. However, protein coding variants affect affinity binding in antibody- or aptamer-based proteomics approaches and may lead to false positive associations. Mass-spectrometry-based (MS) proteomics in principle is able to address this issue by directly identifying the variant protein residues, but the technology faces other challenges, such as exact protein quantification based on multiple peptide measures and library construction accounting for genetic variation while avoiding a combinatoric explosion of library size, which can lead to false positive identifications. Here we investigate these issues by analyzing blood plasma samples from the multi-ethnic Qatar Metabolomics Study on Diabetes (QMDiab) using the Seer Proteograph^TM MS-based proteomics platform. We identified genetic variants that associated at a Bonferroni-level of significance with the detection of matching variant peptides. Interestingly, some of these variants have a matching trans-pQTLs in the much larger UK Biobank OLINK study. They thereby provide experimental evidence for the causal variants and proteins of these associations, confirming expected and suggesting new protein-protein interactions. Our study outlines a novel approach to analyzing MS-proteomics data in the context of genetic association studies and provides new insights into challenges faced by the technology.