Gene-based segregation method for identifying rare variants in family-based sequencing studies

Area of Science:

• Genetics and Genomics
• Statistical Genetics
• Bioinformatics

Background:

• Whole-exome sequencing (WES) in families aids in identifying rare variants for Mendelian and complex diseases.
• Current variant filtering methods lack robust statistical frameworks for family-based association studies.
• Need for advanced statistical approaches to analyze rare variants and their segregation within pedigrees.

Purpose of the Study:

• To introduce a novel gene-based segregation test (GESE) to quantify uncertainty in variant filtering.
• To develop a weighted GESE (wGESE) incorporating additional phenotypes for enhanced statistical power.
• To evaluate the performance of GESE and wGESE against existing region-based methods.

Main Methods:

• Developed GESE based on the probability of segregation events under Mendelian transmission.
• GESE accounts for family relatedness, number of rare functional variants, and minor allele frequencies.
• Implemented a weighted version (wGESE) to integrate supplementary phenotypic data.

Main Results:

• Simulations demonstrate that GESE and wGESE maintain appropriate type I error rates.
• GESE and wGESE exhibit superior statistical power compared to several common region-based methods.
• Application to the Boston Early-Onset COPD study (BEOCOPD) identified promising candidate genes.

Conclusions:

• GESE and wGESE offer a statistically rigorous approach for analyzing family-based sequencing data.
• These methods show significant potential for identifying high-penetrance, large-effect rare coding variants.
• An R package for GESE is available on CRAN, facilitating broader application in genetic research.