FHSA-SED: Two-Locus Model Detection for Genome-Wide Association Study with Harmony Search Algorithm.

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Genome-Wide Association Studies (GWAS) are crucial for identifying disease-related genetic variations.
• Two-locus models are significant but computationally challenging to detect in GWAS.
• Existing methods for two-locus model detection suffer from low power, high computational cost, and model-type bias.

Purpose of the Study:

• To develop a more powerful and computationally efficient method for identifying two-locus disease models in GWAS.
• To overcome the limitations of existing methods, including low detection power and preference for specific disease model types.
• To improve the accuracy and reliability of genetic association analysis.

Main Methods:

• Utilized Bayesian network-based K2-score and Gini-score for characterizing two-locus SNP models.
• Developed an improved Harmony Search Algorithm (HSA) with a local search and two-dimensional tabu table for efficient model searching.
• Employed G-test statistic for robust validation of candidate two-locus models.

Main Results:

• The proposed FHSA-SED method demonstrated superior performance compared to MACOED and CSE in simulations.
• FHSA-SED achieved higher detection power, reduced computation time, and improved evaluation efficiency.
• The method identified two potentially disease-associated SNPs (rs3775652 and rs10511467) in a real Age-related Macular Degeneration (AMD) dataset.

Conclusions:

• FHSA-SED offers a significant advancement in detecting two-locus disease models within GWAS.
• The method provides a robust and efficient approach, enhancing the identification of complex genetic associations.
• FHSA-SED has the potential to accelerate genetic discovery for various complex diseases.