PhenomeXcan: Mapping the genome to the phenome through the transcriptome

Area of Science:

• Genomics
• Transcriptomics
• Translational Bioinformatics

Background:

• Genomic and transcriptomic studies offer insights into complex traits but lack clinical translation due to limited biological context.
• Existing analytic resources are insufficient for linking genetic variants to phenotypic traits effectively.

Purpose of the Study:

• To develop PhenomeXcan, a queryable, gene-based platform integrating genome-wide association study (GWAS) summary statistics with transcriptomic data.
• To prioritize likely causal gene-trait associations using a novel Bayesian colocalization method.

Main Methods:

• Synthesized 8.87 million variants from GWAS summary statistics across 4091 traits with Genotype-Tissue Expression (GTEx) v8 transcriptomic data.
• Developed and applied fast enrichment estimation aided colocalization analysis (fastENLOC) for causal inference.
• Integrated PhenomeXcan with Online Mendelian Inheritance in Man (OMIM) and ClinVar for validation and further analysis.

Main Results:

• Created a comprehensive gene-based platform (PhenomeXcan) covering 22,515 genes and 49 tissues.
• Successfully replicated known gene-trait associations from the phenome-wide association studies (PheWAS) catalog and curated gene lists.
• Identified novel genome-to-phenome associations, complex gene-trait clusters, and shared causal genes between common and rare diseases.

Conclusions:

• PhenomeXcan provides a user-friendly, queryable resource for translational researchers.
• The platform facilitates the discovery of novel gene-trait links and potential therapeutic targets.
• This integrated approach enhances the biological context of genetic associations for clinical applications.