Transmission and decorrelation methods for detecting rare variants using sequencing data from related individuals

Area of Science:

• Genetics
• Statistical Genetics
• Genomic Medicine

Background:

• Whole genome sequencing (WGS) advances enable rare variant investigation, potentially explaining missing heritability.
• Existing rare variant association methods often focus on unrelated individuals, limiting power in family studies.
• Family-based approaches offer enhanced power for detecting rare variant associations.

Purpose of the Study:

• To compare the performance of novel and established family-based methods for rare variant association testing.
• To evaluate methods using simulated whole genome sequencing data from the Genetic Analysis Workshop 19 (GAW19).

Main Methods:

• Comparison of family-based association test for rare variants (FBAT-RV), Minimum p value Optimized Nuisance parameter Score Test Extended to Relatives (MONSTER), family-based association test linear combination (FBAT-LC), and FBAT-Min P.
• Utilized simulated phenotypes and whole genome sequencing data from GAW19.
• Assessed methods based on their power to detect associations between rare/common variants and a quantitative trait.

Main Results:

• MONSTER demonstrated significantly higher overall power compared to FBAT-RV and FBAT-Min P.
• FBAT-LC exhibited comparable overall power to MONSTER.
• MONSTER and FBAT-LC showed the highest power for genes influencing moderate phenotypic variance, while FBAT-LC excelled for genes with the least variance.

Conclusions:

• MONSTER and FBAT-LC are the most powerful methods for rare variant association analysis in related individuals based on GAW19 simulated data.
• Limitations exist for each method, necessitating careful consideration during analysis.
• Future research should aim to integrate the strengths of different methods into a unified family-based rare variant association test.