Higher eQTL power reveals signals that boost GWAS colocalization

Area of Science:

• Genetics
• Genomics
• Molecular Biology

Background:

• Expression quantitative trait locus (eQTL) studies link genetic variants to gene expression.
• eQTLs are proposed to mediate genetic liability for complex traits identified via genome-wide association studies (GWAS).
• Current eQTL studies often lack sufficient power to detect distal regulatory signals due to small sample sizes.

Purpose of the Study:

• To investigate the impact of sample size on eQTL detection power and the characteristics of detected signals.
• To assess how statistical power influences the colocalization of eQTL and GWAS signals.
• To provide guidance for designing future eQTL studies to enhance power and mechanistic insights.

Main Methods:

• Integration of evidence from multiple eQTL studies.
• Statistical analysis of eQTL detection rates across varying signal strengths and sample sizes.
• Comparison of characteristics (e.g., distance to gene, loss intolerance) of eQTLs detected in different sample sizes.
• Meta-analysis of eQTL and GWAS colocalization.

Main Results:

• Limited statistical power in smaller sample sizes (<2,000 individuals) significantly skews eQTL detection, missing many true signals.
• A sample size of 2,000 is estimated to detect only 36.8% of all eQTLs.
• eQTLs detectable only in larger studies share characteristics with GWAS signals, such as greater distance to the target gene.
• Increased sample sizes in meta-analyses led to a substantial rise in detected eQTL-GWAS colocalizations.

Conclusions:

• Underpowered eQTL studies may lead to underestimation of the role of gene expression in complex traits.
• Larger sample sizes are essential for comprehensive eQTL discovery and robust identification of gene-trait mechanisms.
• Future eQTL study designs should prioritize increased sample sizes to improve power and complement perturbation-based methods.