Robust Rare-Variant Association Tests For Quantitative Traits in General Pedigrees.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Next-generation sequencing enables identification of rare genetic variations linked to complex traits.
• Existing statistical methods primarily target case-control or population studies, with limited applicability to family-based designs.
• Current family-based methods are often restricted to small family units (trios, nuclear families), lacking capacity for large, complex pedigrees.

Purpose of the Study:

• To develop a novel statistical method for rare-variant analysis in large pedigree studies.
• To leverage information from all available relatives within a pedigree structure.
• To address the limitations of existing methods for complex family data.

Main Methods:

• A kernel-machine regression (KMR) framework is employed for its high power and efficient p-value calculation.
• Integration with a Quantitative Transmission/Disequilibrium Test (QTDT) framework ensures robustness against population stratification.
• A two-component approach calculates expected genotypes (between-family) and genotype deviates (within-family), with the latter being stratification-robust.

Main Results:

• The proposed method effectively analyzes rare variants in large pedigrees.
• The kernel-machine regression approach demonstrates high statistical power.
• The integrated QTDT framework successfully mitigates population stratification effects.

Conclusions:

• The developed method fills a critical gap in rare-variant analysis for large, complex family studies.
• This approach offers a powerful and robust tool for genetic association studies in diverse family structures.
• The method's applicability is demonstrated through simulations and real-world genetic data analysis.