Gene-based association tests in family samples using GWAS summary statistics.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Genome-wide association studies (GWAS) identify genetic variants linked to phenotypes using summary statistics.
• Current gene-based tests require individual-level data or are limited to unrelated individuals.
• Linear mixed models (LMMs) are effective for GWAS in family samples but pose challenges for summary statistic-based gene tests.

Purpose of the Study:

• To develop a method for gene-based association tests using GWAS summary statistics from family samples analyzed with LMMs.
• To adapt existing summary statistic-based methods for unrelated individuals to family data.

Main Methods:

• Approximating the correlation matrix of marginal Z-scores using the linkage disequilibrium matrix.
• Leveraging the diagonal block structure of the kinship matrix in LMMs.
• Applying existing gene-based association test methods for unrelated individuals directly to family data.

Main Results:

• The proposed strategy effectively controls the type 1 error rate in simulations across various scenarios.
• The method allows the direct application of current summary statistic-based tests to family data.
• The approach was successfully demonstrated on a dental caries GWAS dataset.

Conclusions:

• Gene-based association tests can be performed on family samples using GWAS summary statistics from LMMs.
• The developed method expands the utility of publicly available GWAS summary statistics for family-based genetic studies.
• This approach facilitates broader genetic discovery in family cohorts.