Empirical-Bayesian Testing for Family Genome-wide Association Data

描述（由申请人提供）：随着新的基因分型技术的出现，特别是SNP阵列以及即将到来的下一代测序，人类疾病/性状的基因定位工作已经集中在遗传关联研究上。基于家庭的研究和基于人群的研究是遗传关联研究的两种常用设计。与以人群为基础的研究相比，以家庭为基础的研究对因人群分层而产生的偏倚具有稳健性。然而，基于家庭的研究往往不如基于人口的研究有力，因为家庭之间的关联信息由于易受人口分层偏差的影响而未被使用。本应用程序旨在开发一种新的分析工具，以充分利用数据并保持对人口分层的稳健性。具体而言，我们提出了一种基于经验贝叶斯框架的统计方法，该方法估计个体基因座的群体分层偏差，使家庭间的信息量在偏差的基础上缩小。为了证明与现有方法相比，所提出方法的有效性和优越性能，我们计划通过广泛的模拟来评估它，并将其应用于dbGaP中的家族全基因组关联数据。

DESCRIPTION (provided by applicant): With availability of new genotyping technology, in particular SNP arrays as well as the coming next generation sequencing, efforts of mapping genes of human diseases/traits have been focusing on genetic association study. Family-based and population-based studies are two commonly-used designs of genetic association study. In contrast to population-based study, family based study is robust to bias due to population stratification. However, family-based study is often less powerful than population-based study because association information between families is not used due to its susceptibility to the bias of population stratification. The aim of this application is to develop a new analytic tool to fully utilize data as well as maintain the robustness to population stratification. Specifically, we propose a new statistic approach under the Empirical Bayesian framework, in which the bias of population stratification of individual loci is estimated so that the amount of information between families contributed to the testing statistic shrinks based on the bias. To demonstrate the validity and superior performance of the proposed approach compared to approaches available, we plan to evaluate it by extensive simulations and apply it to the family genome-wide association data in dbGaP. PUBLIC HEALTH RELEVANCE: With availability of new genotyping technologies, efforts of mapping genes of human diseases/traits have been focusing on genome-wide association study (GWAS). Family-based and population-based studies are two commonly-used designs used in GWAS, and each of them has unique advantages and disadvantages. The aim of this application is to develop a new analytic tool to make use of advantages of both designs - to maintain the robustness against population stratification and to achieve higher power. The proposed study has the potential to facilitate the research for identifying novel loci related to complex diseases.