Omnibus testing approach for gene-based gene-gene interaction.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Complex traits are influenced by genetic interactions, a key heritable component.
• Genome-wide association studies (GWAS) have enabled SNP-level interaction analysis, but gene-based approaches offer improved power and interpretation.
• Existing gene-based interaction methods struggle with the vastness of interaction patterns due to multidimensional modeling.

Purpose of the Study:

• To develop a robust gene-based gene-gene interaction testing approach.
• To address the limitations of existing methods in handling complex interaction patterns.
• To propose an omnibus test leveraging heterogeneity and coding complementarity for enhanced detection power.

Main Methods:

• Utilized a logistic regression modeling framework to combine SNP-SNP interaction tests.
• Developed an omnibus test integrating existing global tests and complementary SNP coding strategies (allele-based and genotype-based).
• Conducted extensive simulations to evaluate the test's performance across various genetic models.

Main Results:

• The proposed omnibus test demonstrated high power in detecting common and complex genetic interaction models, including those with multiple causal pairs.
• The approach showed robustness and improved power compared to single global tests in simulation and replication studies.
• Application to real datasets confirmed the method's adaptability in replicating gene-gene interactions.

Conclusions:

• The developed omnibus test provides a powerful and flexible tool for gene-based gene-gene interaction analysis.
• This method enhances the detection of genetic interactions underlying complex traits.
• The approach is adaptable for real-world genetic studies and replication efforts.