Finally, we assessed the contribution of eQTLs & sQTLs to GWAS risk loci, finding colocalizations both shared and unique to gene expression and splicing. Those overlapping proteomic or metabolic traits revealed potential pathways through which risk loci may be acting. (7/7)

Using this colocalized subset, we performed mediation analysis using individuals with overlapping multi-omic traits to interrogate the strength and direction of shared genetic effects, observing 222 molecular phenotypes significantly mediated by gene expression or splicing. (6/7)

Using multi-omic data collected from the same individuals and large external cohorts, we performed colocalization analysis to identify eQTLs and sQTLs also regulating proteomic or metabolomic traits, observing a total of 3,430 with a shared association signal. (5/7)

We additionally mapped trans-eQTLs and -sQTLs acting through cis-eGenes using the set of independent cis-eQTL variants. With this mechanism, we identified 2,058 trans-eGenes, and 644 splicing events (in 209 trans-sGenes), the most through the RNA-binding splice-factor QKI. (4/7)

Linking expression and splicing phenotypes to genotype, we discovered cis-QTLs for 17,233 eGenes, and 29,514 splicing phenotypes (in 6,853 sGenes). (3/7)

This was a huge effort by the whole team and particularly my co-first author Elodie Persyn, integrating proteomic and metabolite QTL data from the same individuals, and external health GWAS data to assess the genetic contribution of eQTLs & sQTLs to downstream traits. (2/7)