Quantile Rank Test for Dynamic Heterogeneous Genetic Effects in Longitudinal Electronic Health Record Analysis

• Activity Number: 43 - Statistical Genetics II – New Models for Complex Study Designs
• Type: Contributed
• Date/Time: Monday, August 3, 2020 : 10:00 AM to 2:00 PM
• Sponsor: Section on Statistics in Genomics and Genetics
• Author(s): Tianying Wang* and Iuliana Ionita-laza and Ying Wei and Chunhua Weng and Zixu Wang
• Companies: Columbia University and Columbia University and Columbia University and Columbia University and Columbia University
• Keywords: quantile regression; genetic association; electronic health records; heterogeneous associations

Abstract

Over the past few years, an increasing number of sequence-based association studies evaluated the group-wise effects of rare and common genetic variants and identified significant associations between a gene and a phenotype of interest. Utilizing the longitudinal trajectory of outcomes can help us explore the dynamic genetic effect and improve the test power. However, limited researches have been done in such area especially for electronic health records (EHR) data. In this paper, we propose a generalized integrated rank score test based on quantile regression, which considering the quantile effect of the entire sample. A scalable fast algorithm is provided for large EHR data, and a perturbation method is used to overcome the deflated/inflated type I error issue for longitudinal studies. We will first introduce the test without longitudinal data, then we show its generalization and application results in the Electronic Medical Records and Genomics (eMERGE) Network data.