A forest-based approach to identifying gene and gene gene interactions

Area of Science:

• Genetics
• Computational Biology
• Bioinformatics

Background:

• Complex diseases often result from multiple interacting genes and environmental factors.
• Identifying these gene-gene and gene-environment interactions remains a significant challenge in genetic research.
• Existing methods struggle to effectively pinpoint these complex interactions.

Purpose of the Study:

• To develop and validate a novel forest-based approach for identifying gene-gene and gene-environment interactions.
• To introduce a concept of variable importance for assessing the contribution of genetic factors.
• To apply the method to real-world data for disease risk identification.

Main Methods:

• A forest-based computational approach was developed, utilizing a concept of variable importance.
• The method's validity was assessed through simulation studies.
• The approach was applied to a real dataset concerning age-related macular degeneration (AMD).

Main Results:

• The proposed method effectively identified gene-gene and gene-environment interactions in both simulated and real data.
• Analysis of AMD data confirmed a known genetic variant (P = 2E-6).
• A novel, potentially protective haplotype near SNP rs10272438 on chromosome 7 was identified (P = 0.0024) after accounting for numerous SNPs.

Conclusions:

• The developed forest-based method is a powerful and flexible tool for identifying high-risk haplotypes and their interactions.
• The study identified a new potentially protective genetic variant for age-related macular degeneration (AMD).
• This approach advances the ability to unravel the genetic architecture of complex diseases.