Testing for direct genetic effects using a screening step in family-based association studies.

Area of Science:

• Genetics
• Biostatistics
• Epidemiology

Background:

• Family-based genome-wide association studies (GWAS) often exhibit lower statistical power compared to population-based designs.
• Assessing direct genetic effects on phenotypes, independent of secondary traits, is challenging in family studies.
• Multiple single nucleotide polymorphism (SNP) testing in family-based GWAS necessitates methods to manage the burden of multiple comparisons.

Purpose of the Study:

• To propose a novel two-stage screening method for family-based association tests (FBAT).
• To integrate this screening step into existing causal analysis frameworks, such as the Van Steen algorithm.
• To enhance the detection of direct genetic effects in family-based GWAS.

Main Methods:

• A two-stage screening procedure is introduced for family-based association tests (FBAT).
• The proposed method is integrated into a framework similar to the conditional mean model approach.
• The method's performance is evaluated using simulation studies and applied to real-world data.

Main Results:

• Simulations confirm that the proposed screening method effectively preserves the type 1 error rate.
• The method demonstrates advantages when analyzing multiple genetic markers simultaneously.
• The application to the Framingham Heart Study illustrates its practical utility.

Conclusions:

• The developed two-stage screening method enhances the analysis of family-based GWAS.
• This approach offers a robust way to minimize multiple comparison issues in causal analyses.
• The method provides a valuable tool for identifying direct genetic associations in family studies.