A novel method to detect gene-gene interactions in structured populations: MDR-SP

Area of Science:

• Genetics and Genomics
• Statistical Genetics
• Population Genetics

Background:

• Complex diseases involve multiple genes and environmental factors, necessitating the study of gene-gene and gene-environment interactions.
• Traditional statistical methods struggle with high-order interactions and sparse data, and may miss interactions lacking detectable main effects.
• Existing methods for genetic association studies often fail to simultaneously detect interactions and control for population stratification.

Purpose of the Study:

• To develop a novel statistical test for detecting gene interactions in the presence of population stratification.
• To address the limitations of existing methods, particularly the multifactor dimensionality reduction (MDR) method, in handling population structure.
• To create a method applicable to both quantitative and qualitative traits, incorporating covariates.

Main Methods:

• Development of a new statistical test named MDR in structured populations (MDR-SP).
• MDR-SP is designed to detect high-order gene-gene and gene-environment interactions.
• The method incorporates a mechanism to control for population stratification, a common confounder in genetic studies.

Main Results:

• Simulation studies were conducted to validate the MDR-SP test.
• The simulations demonstrated the test's ability to detect interactions effectively.
• The power of the MDR-SP test was evaluated under various scenarios.

Conclusions:

• MDR-SP is a valid and powerful statistical tool for detecting gene interactions in structured populations.
• This method enhances the accuracy of genetic association studies by simultaneously addressing interactions and population stratification.
• MDR-SP offers a significant advancement for studying the genetic basis of complex diseases.