A statistical method for rare variants association studies in pedigree data

Objective: To propose an adjusted sequence kernel association test (SKAT) model in order to identify rare variants for pedigree data which has higher statistical power. Methods: In this paper, we proposed a SKAT model fitting pedigree data (ADSKAT). The SKAT model was modified by adding a random effect vector of pedigree structure into the model. Thus the influence of kinship correlation was taken into consideration in the new model. A new distribution of test statistics was defined. Results: Simulations demonstrated that ADSKAT well controlled the inflation of type I error and achieved better statistical power than the existed mainstream methods for identifying disease-related rare variants. Conclusions: ADSKAT has broad application prospects in the fields of identifying disease related rare variants in pedigree data.