An "almost exhaustive" search-based sequential permutation method for detecting epistasis in disease association...

Area of Science:

• Genetics
• Computational Biology
• Disease Etiology

Background:

• Complex diseases often result from interactions between multiple genes.
• Identifying "uncommon but strong" (UBS) genetic effects, where rare allelic combinations confer high disease risk, is challenging with standard association tests.

Purpose of the Study:

• To develop and validate a novel computational method for the powerful detection of UBS genetic interaction effects.
• To improve the identification of complex disease etiological factors.

Main Methods:

• A two-module method combining a pattern counting algorithm for risk evaluation and a sequential permutation scheme for multiple testing correction.
• Demonstration using a candidate gene dataset for cardiovascular diseases with an introduced UBS three-locus interaction.
• Simulation studies using a joint dominance three-locus model to assess power and false rejection rates.

Main Results:

• The proposed method demonstrates significantly higher power in detecting UBS interactions compared to standard approaches like the trend test and multifactor dimensionality reduction.
• Effective identification of injected UBS three-locus interactions in cardiovascular disease candidate gene data.
• Simulation results confirm the method's robust performance in identifying UBS effects.

Conclusions:

• The developed method offers a powerful new tool for uncovering complex genetic interactions underlying common diseases.
• This approach enhances our ability to identify high-risk genetic profiles, advancing the study of disease etiology.