AprioriGWAS, a new pattern mining strategy for detecting genetic variants associated with disease through interaction...

Area of Science:

• Genetics
• Bioinformatics
• Statistical Genetics

Background:

• Identifying gene-gene interactions is crucial for understanding complex diseases.
• Genome-wide association studies (GWAS) face challenges in efficiently searching vast combinations of genetic variants and testing for epistasis.
• Existing methods struggle with the computational complexity of exploring all possible genotype combinations.

Purpose of the Study:

• To develop an efficient and accurate method for identifying gene-gene interactions in GWAS.
• To address the challenges of variant combination selection and epistasis testing.
• To apply the developed method to identify novel genetic interactions in age-related macular degeneration (AMD) and bipolar disorder.

Main Methods:

• Developed AprioriGWAS, leveraging Frequent Itemset Mining (FIM) for efficient genotype pattern identification.
• Implemented a conditional permutation procedure for reliable statistical inference of epistasis using Pearson's chi-square test.
• Applied AprioriGWAS to AMD and bipolar disorder datasets.

Main Results:

• AprioriGWAS successfully identified significant gene-gene interactions in both AMD and bipolar disorder datasets.
• In AMD, interactions between ANGPT1, retinal genes, and CFH were found, with enrichment in "glycosaminoglycan biosynthetic process".
• In bipolar disorder, significant interactions were detected for variants without marginal effects, including within GABRB2 and GRIA1 genes.

Conclusions:

• AprioriGWAS is an effective tool for discovering complex gene-gene interactions in GWAS.
• The identified interactions in AMD and bipolar disorder provide new insights into disease mechanisms.
• The method highlights the importance of exploring non-marginal genetic effects for understanding disease etiology.