A note on p-value multiple-testing adjustment for two-step genome-wide gene-environment interactions scans.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Genome-wide interaction scans (GWIS) are crucial for identifying genetic interactions.
• Two-step testing is a computationally efficient state-of-the-art approach for GWIS, offering higher power than single-step methods.
• A key limitation of two-step tests is the absence of valid p-values for direct comparison with single-step results, hindering interpretation.

Approach:

• This research defines multiple-testing adjusted p-values for two-step tests.
• The study leverages standard multiple-testing theory to derive these p-values.
• A novel scaling method is introduced to ensure valid comparisons between two-step and single-step test results.

Key Points:

• Two-step tests control the genome-wide type I error rate effectively.
• The proposed method generates valid, adjusted p-values for two-step GWIS.
• These adjusted p-values can be scaled for direct comparison with single-step GWIS results.

Conclusions:

• The developed methodology addresses the p-value comparability issue in two-step GWIS.
• This facilitates more straightforward interpretation and integration of results from different GWIS approaches.
• The findings enhance the utility of two-step testing in genetic research.