| Summary: | 碩士 === 國立陽明大學 === 公共衛生研究所 === 106 === FGESKAT (Family-Based Gene-Environment Interaction Using Sequence Kernel Association Test) is a family-based sequence kernel association test, which tests for the joint effect of gene variants (common and rare) and gene-environment interaction while easily adjusting for covariates. Such kernel score statistic allows for familial dependencies and adjusts for random confounding effects. However, the adjustment for p-value was too conservative to obtain the justifiable results. Therefore, this thesis derives the optimal test for FGESKAT, calculates p-value using Monte Carlo methods. The new strategy was applied to whole genome sequence data in Genetic Analysis Workshop 18 (GAW18) and discovered concordance and discordant regions comparing to methods without interactions. Optimal test for FGESKAT identified significant results relate to the cardiovascular diseases that can be replicated, including HDAC9 and CACNA2D1. Our results show that the power of optimal test have been highly improved.
|
|---|
