Family-based Association Test Using Both Common and Rare Variants and Accounting for Directions of Effects

Abstract

Association studies aim to find genetic variants responsible for complex traits, such as human diseases. Next-generation sequencing (NGS) has generated many rare variants for association analysis. As a complex disease is caused by multiple genetic variants, while the effect for each of the variants on the disease is not large, how to identify the group of disease susceptibility variants is challenging. Moreover, the complex disease can be caused by the joint effects of common and rare variants. Statistical tests that can accommodate both common and rare variants become important. Taking advantages of some properties of family-based studies, we developed the Variable Threshold Pedigree Disequilibrium Test (VT-PDT) for family sequence data. VT-PDT is developed based on the family-based Pedigree Disequilibrium Test (PDT). The variable threshold algorithm is used in VT-PDT to computationally search for an optimal allele frequency threshold that rare variants below the threshold are more likely to be functional for the disease. A normalized log relative risk for each of the common variants calculated based on parental mating types is used to select common variants that are more likely to be associated with the disease. The statistics from both rare and common variants are combined as the VT-PDT statistic. A permutation procedure is used in VT-PDT to approximate the distribution of VT-PDT and calculate the p-value. We used realistic simulations and examples to demonstrate that VT-PDT is a powerful test for family-based association analysis for sequence data.?