An Efficient Multiple-Testing Adjustment for eQTL Studies that Accounts for Linkage Disequilibrium between Variants.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Multiple-testing correction in cis-eQTL studies balances statistical power and computational efficiency.
• Traditional methods like Bonferroni correction are overly conservative, while permutation tests are computationally intensive.
• Existing methods struggle to account for linkage disequilibrium between genetic variants.

Purpose of the Study:

• To introduce eigenMT, a novel and computationally efficient method for multiple-testing correction in cis-eQTL studies.
• To provide adjusted p-values that closely approximate empirical results from computationally expensive permutation tests.
• To enhance the efficiency of eQTL discovery pipelines across multiple tissues or conditions.

Main Methods:

• Developed eigenMT, a method estimating the effective number of independent variants (Meff) using eigenvalue decomposition of the genotype correlation matrix.
• Utilized a regularized estimator for the correlation matrix to ensure robustness and accuracy of Meff.
• Applied eigenMT to genotype data, demonstrating its efficiency for multi-tissue or multi-condition eQTL studies.

Main Results:

• eigenMT runs over 500 times faster than traditional permutation-based methods.
• Adjusted p-values generated by eigenMT closely approximate empirical p-values.
• Demonstrated increased efficiency when applying eigenMT across multiple tissues or conditions using a common genotype matrix.

Conclusions:

• eigenMT offers a simpler, more efficient, and accurate alternative for multiple-testing correction in cis-eQTL analyses.
• The method integrates seamlessly into existing eQTL discovery pipelines.
• eigenMT improves the balance between statistical power and computational cost, facilitating broader eQTL research.