ProGeM

ProGeM prioritizes candidate causal genes (mediating genes) at molecular quantitative trait loci (QTLs) by integrating biological-domain annotations and genome annotation to identify genes underlying molecular QTLs observed in GWAS of metabolites, lipids, and proteins.

Key Features:

• Candidate gene prioritization: Prioritizes candidate causal (mediating) genes associated with molecular QTLs identified in genome-wide association studies (GWAS).
• Annotation integration: Integrates biological domain-specific annotation data with comprehensive genome annotation from multiple repositories.
• Validation on metabolite QTLs: Benchmarked using a reference set of 227 well-curated metabolite QTLs, with the expert-assigned causal gene included among prioritized genes at 98% of loci.
• Genomic proximity analysis: Demonstrates that 69% of true positive causal genes are nearest to the sentinel variant at molecular QTLs.
• cis-eQTL benchmarking: Shows that relying on cis-gene expression QTL data produces a higher false-positive rate, with three incorrect assignments for every correct one.
• Applicability to molecular phenotypes: Applicable to molecular phenotypes including metabolites, lipids, and proteins.

Scientific Applications:

• Prioritization in GWAS: Prioritizes mediating genes underlying molecular QTLs from GWAS of metabolites, lipids, and proteins.
• Molecular QTL annotation: Assigns candidate causal genes to molecular QTLs to link genetic variants to potential mediating genes.
• Evaluation of prioritization criteria: Enables comparative assessment of genomic proximity versus cis-eQTL evidence to inform gene-prioritization strategies and reduce false positives.

Methodology:

Integrates biological domain-specific annotation data with comprehensive genome annotation from multiple repositories, benchmarks performance against a reference set of 227 metabolite QTLs, and analyzes nearest-gene to sentinel variant and comparisons with cis-gene expression QTL data.