Sequence Kernel Association Test for Multivariate Quantitative Phenotype in Family Samples. Q. Yan1, B. Li2, W. Chen1, N. Liu3 1) Division of Pulmonary Medicine, Allergy and Immunology, Children's Hospital of Pittsburgh of UPMC, Pittsburgh, PA, 15224, USA; 2) Department of Molecular Physiology & Biophysics, and Neurology, Vanderbilt University Medical Center, Nashville, TN 37232, USA; 3) Department of Biostatistics, University of Alabama at Birmingham, Birmingham, AL, 35294, USA.
The recent development of sequencing technology allows identification of the association between rare genetic variants and complex diseases. Over the past few years, a number of rare variant association test approaches have been developed. Among these methods, the kernel machine test as a set-based method has been shown to perform well in different scenarios. Many genetic studies have been conducted to collect multiple correlated phenotypes for one complex disease. Because of the relatedness between phenotypes, jointly testing the association between phenotypes and genetic variants may increase the power to detect causal genes. In addition, family based designs have been widely used to study the association between diseases and genetic factors. Thus, familial correlation needs to be appropriately handled to avoid inflated type I error rate. In this work, we aim to conduct the association test of rare variants in family samples, which uses multiple phenotype measurements for each subject. Our proposed method uses kernel machine regression and denoted as MF-SKAT. It is based on linear mixed model framework and can be applied to a larger range of studies with different types of traits. In our simulation studies, the results show that the kernel machine test (M-SKAT considering the correlation between multiple phenotypes) has inflated Type I error rate when applying to familial data directly. By contrast, our proposed MF-SKAT has correct Type I error rate. Furthermore, MF-SKAT jointly analyzing phenotypes has increased power comparing to the methods separately analyzing phenotypes (F-SKAT considering the family structure) in all the scenarios we considered. Finally, we illustrate our proposed methodology by analyzing whole-genome genotyping data from a study of lung function and exome sequencing data from a study of Anorexia Nervosa.
You may contact the first author (during and after the meeting) at