DualWMDR: Detecting epistatic interaction with dual screening and multifactor dimensionality reduction

Area of Science:

• Genetics and Genomics
• Computational Biology
• Biostatistics

Background:

• Identifying genetic susceptibility for complex diseases often involves detecting epistatic interactions.
• Multifactor dimensionality reduction (MDR) is a common approach for epistasis detection, but existing methods face challenges with high computational costs and suboptimal performance.
• There is a need for more efficient and accurate methods to identify gene-gene interactions contributing to complex diseases.

Purpose of the Study:

• To propose a novel method, DualWMDR, that integrates a dual screening strategy with MDR to improve epistasis detection.
• To enhance the accuracy and computational efficiency of identifying genetic factors underlying complex diseases.
• To evaluate the performance of DualWMDR against existing methods using simulation and real-world genomic datasets.

Main Methods:

• Developed DualWMDR, a novel approach combining a dual screening strategy with MDR for epistasis detection.
• The first screening step utilizes adaptive clustering and part mutual information (PMI) to group single nucleotide polymorphisms (SNPs) and filter noisy data.
• The second screening step selects dominant SNPs by considering both single-locus and interaction effects, creating a refined candidate set for MDR, which then employs weighted classification for improved performance.

Main Results:

• DualWMDR demonstrated superior performance compared to existing competitive methods on various simulation datasets.
• The method's effectiveness was further validated on three real genome-wide datasets: age-related macular degeneration (AMD), breast cancer (BC), and celiac disease (CD).
• Results consistently showed DualWMDR's capability in accurate epistasis identification.

Conclusions:

• DualWMDR offers a significant advancement in detecting epistatic interactions for complex diseases.
• The proposed dual screening strategy effectively reduces computational burden and enhances the accuracy of MDR.
• DualWMDR provides a robust and efficient tool for genetic susceptibility studies, applicable to diverse complex diseases.