ProGeM: a framework for the prioritization of candidate causal genes at molecular quantitative trait loci

Abstract

Quantitative trait locus (QTL) mapping of molecular phenotypes such as metabolites, lipids and proteins through genome-wide association studies represents a powerful means of highlighting molecular mechanisms relevant to human diseases. However, a major challenge of this approach is to identify the causal gene(s) at the observed QTLs. Here, we present a framework for the 'Prioritization of candidate causal Genes at Molecular QTLs' (ProGeM), which incorporates biological domain-specific annotation data alongside genome annotation data from multiple repositories. We assessed the performance of ProGeM using a reference set of 227 previously reported and extensively curated metabolite QTLs. For 98% of these loci, the expert-curated gene was one of the candidate causal genes prioritized by ProGeM. Benchmarking analyses revealed that 69% of the causal candidates were nearest to the sentinel variant at the investigated molecular QTLs, indicating that genomic proximity is the most reliable indicator of 'true positive' causal genes. In contrast, cis-gene expression QTL data led to three false positive candidate causal gene assignments for every one true positive assignment. We provide evidence that these conclusions also apply to other molecular phenotypes, suggesting that ProGeM is a powerful and versatile tool for annotating molecular QTLs. ProGeM is freely available via GitHub.

A multi-tissue transcriptome analysis of human metabolites guides the interpretability of associations based on multi-SNP models for gene expression

“…These included two genes involved in uridine metabolism - CDA and UPP1 . Notably, SNPs in at these two loci were sub-genome- wide significant in the GWAS but were implicated from our S-PrediXcan analysis and subsequent studies 17 ( Table S7) . We expanded the search further by querying a recently curated dataset 17 and found an additional 18 genes annotated to at least one metabolic pathway.…”

“…Notably, SNPs in at these two loci were sub-genome- wide significant in the GWAS but were implicated from our S-PrediXcan analysis and subsequent studies 17 ( Table S7) . We expanded the search further by querying a recently curated dataset 17 and found an additional 18 genes annotated to at least one metabolic pathway. Thus, as many as 37% (114/312 gene-metabolite pairs) of novel TWAS gene associations can be considered biologically plausible albeit based on the rather permissive overlap between “metabolic pathway” and met-QTL.…”

“…We then performed a more stringent evaluation by determining the number of novel gene-metabolite associations (again excluding “bystander” genes) where the novel gene either shared at least one metabolic pathway with a reported truth set gene for the associated metabolite or has been curated as a high confidence causal gene with the associated metabolite in recent publications 17 . We found that 16 (5%) of the 312 novel gene-metabolite pairs met this criterion (Table S8) .…”

“…Moreover, if such models better reflect the number of independent genetic signals acting on a phenotype, are they supported by evidence of shared identity between the trait-associated and eQTL variants within the model? Furthermore, the extent to which novel genes implicated by colocalized associations represent genuine biological relationships, causal for disease, is unclear and inference is further complicated by the shared regulatory architecture of gene expression and by horizontal pleiotropy 16,17 .…”

A robust and efficient method for Mendelian randomization with hundreds of genetic variants: unravelling mechanisms linking HDL-cholesterol and coronary heart disease

“…r-project.org/web/packages/MendelianRandomization/index.html). a Nearest gene is RP11-136O12.2, but signal maps to the TRIB1 gene [32]. b This gene is adjacent to the APOA5 -APOA4 -APOC3 -APOA1 cluster on chromosome 11.…”