Multiple testing correction in linear mixed models.

Area of Science:

• Genetics
• Statistical genomics
• Bioinformatics

Background:

• Genome-wide association studies (GWAS) involve analyzing millions of genetic markers, presenting a significant multiple hypothesis testing challenge.
• Permutation tests are the gold standard for multiple testing correction in GWAS due to their ability to account for genomic correlation.
• Linear mixed models (LMMs) are now standard in GWAS for handling population structure and improving power, but lack applicable multiple testing correction methods.

Purpose of the Study:

• To develop an efficient and accurate multiple testing correction method specifically for linear mixed models (LMMs) used in genome-wide association studies (GWAS).
• To address the computational limitations of existing methods when applied to LMMs in large-scale genetic analyses.

Main Methods:

• Developed a novel approach for estimating per-marker thresholds within the linear mixed model framework.
• Validated the method using both simulated and real-world datasets from human, mouse, and yeast studies.

Main Results:

• The proposed method accurately estimates per-marker thresholds, comparable to the gold-standard permutation test.
• Computation time was drastically reduced from months to hours, demonstrating significant efficiency gains.
• The approach was successfully applied to diverse datasets, confirming its accuracy and broad applicability.

Conclusions:

• An efficient and accurate multiple testing correction approach for LMMs in GWAS has been established.
• The study provides insights into the interplay between per-marker thresholds, genetic relatedness, and heritability in genetic analyses.