Genotype Pattern Mining for Pairs of Interacting Variants Underlying Digenic Traits

Area of Science:

• Genetics
• Computational Biology
• Bioinformatics

Background:

• Digenic traits result from the interaction of two DNA variants, complicating genetic disease diagnosis.
• Standard genetic methods struggle to detect digenic traits, especially when individual variants have minor effects.

Purpose of the Study:

• To develop and evaluate a frequent pattern mining (FPM) method for identifying pairs of genotypes associated with digenic traits.
• To improve the detection of complex genetic interactions in case-control studies.

Main Methods:

• Applied frequent pattern mining (FPM) to identify genotype patterns of length two.
• Utilized permutation testing to assign p-values for genotype patterns, testing the null hypothesis of equal frequencies in cases and controls.
• Compared FPM with other interaction search approaches using published datasets.

Main Results:

• The FPM approach effectively identifies pairs of genotypes that discriminate between cases and controls.
• Demonstrated the utility of FPM for detecting digenic traits where individual variants have minimal impact.
• Provided a freely available implementation of the FPM method for case-control studies.

Conclusions:

• Frequent pattern mining offers a powerful computational strategy for uncovering digenic traits.
• This method enhances the ability to detect complex genetic interactions, advancing the understanding of genetic diseases.
• The open-source implementation facilitates broader application in genetic research.