A 2-step penalized regression method for family-based next-generation sequencing association studies

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Large-scale genetic studies frequently involve related individuals, posing computational challenges.
• Existing methods for analyzing familial data can be cumbersome and inefficient.
• Integrating rare and common variants in pedigree analysis requires advanced statistical approaches.

Purpose of the Study:

• To develop an efficient computational method for analyzing sequencing data from complex pedigrees.
• To incorporate both rare and common genetic variants into the analysis of related individuals.
• To provide a robust framework for genetic association studies in multigenerational cohorts.

Main Methods:

• A two-step procedure was developed to manage familial relatedness.
• Familial correlation was sequentially regressed out from the data.
• Penalized regression was applied to phenotypic residuals for variant association testing within genetic units.

Main Results:

• The proposed method efficiently handles complex pedigree structures.
• The approach successfully incorporates information from both rare and common variants.
• Simulation studies demonstrated the operating characteristics of the method using a large multigenerational cohort.

Conclusions:

• The presented approach offers an efficient solution for genetic studies involving complex family structures.
• This method facilitates the integration of diverse genetic variant information (rare and common) in association analyses.
• The findings support the utility of this approach for large-scale genetic epidemiology and the study of complex diseases.