Kullback-Leibler distance methods for detecting disease association with rare variants from sequencing data

Area of Science:

• Genetics
• Bioinformatics
• Statistical genomics

Background:

• Next-generation sequencing enables comprehensive genotyping of genetic variations.
• Investigating rare variants' impact on complex diseases is crucial.
• Existing methods face challenges with rare variants and complex genetic architectures.

Purpose of the Study:

• To propose novel Kullback-Leibler distance-based Tests (KLTs) for identifying genotypic differences between cases and controls.
• To develop a method robust to variants with opposing effects and null variants.
• To create a unified approach for analyzing rare and common variants without arbitrary thresholds.

Main Methods:

• Developed four Kullback-Leibler distance-based Tests (KLTs).
• Explicitly compared genotype distributions to handle variants with opposite directional effects.
• Incorporated a noise-fighting mechanism for robustness against null variants.
• Implicitly accounted for variant correlations and linkage disequilibrium (LD) structure.

Main Results:

• KLTs demonstrated superior performance compared to the sum of squared score test (SSU) and optimal sequence kernel association test (SKAT-O) in simulations.
• The proposed tests effectively analyze rare and common variants together.
• KLTs showed robustness to null variants and handled complex LD structures.
• Successful application to the Dallas Heart Study data confirmed feasibility.

Conclusions:

• Kullback-Leibler distance-based Tests (KLTs) offer a powerful and flexible approach for genetic association studies.
• KLTs provide a robust framework for analyzing complex diseases influenced by rare and common variants.
• The developed methods enhance the ability to detect genetic contributions to complex diseases using next-generation sequencing data.