HiSeeker: Detecting High-Order SNP Interactions Based on Pairwise SNP Combinations.

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Genome-wide association studies (GWAS) aim to explain complex disease heritability.
• Existing methods often overlook high-order single nucleotide polymorphism (SNP) interactions.
• Current high-order interaction detection methods struggle with genome-wide data and low power.

Purpose of the Study:

• To develop a flexible, two-stage approach (HiSeeker) for detecting high-order SNP interactions.
• To address the limitations of existing methods in handling large-scale genomic data and improving detection power.

Main Methods:

• A two-stage approach: screening followed by search.
• Screening stage uses chi-squared tests and logistic regression for candidate pairwise SNP combinations.
• Search stage employs exhaustive search and ant colony optimization for high-order interactions.

Main Results:

• HiSeeker demonstrated superior efficiency and effectiveness in detecting high-order interactions on simulated data compared to existing algorithms.
• Identified significant high-order interactions in real case-control datasets, often missed by other methods.
• These detected interactions involved SNPs with weak individual or pairwise effects.

Conclusions:

• HiSeeker provides a powerful and efficient tool for identifying complex, high-order SNP interactions.
• The approach enhances the ability to explain missing heritability in common complex diseases.
• HiSeeker advances the field of genetic interaction analysis, particularly for large-scale GWAS.