Correcting for Population Structure in Gene-by-Environment Interaction Studies

描述（由申请人提供）：在过去的几年中，全基因组关联研究（GWAS）已经确定了许多与常见人类疾病相关的基因。在这些研究中，收集了数千个个体的遗传变异，并将其与这些个体的疾病状态相关联。GWAS的一个挑战性方面是所收集的个体彼此之间的相关程度不同。这可能导致虚假的关联，这些关联是似乎与疾病相关的基因，但实际上是个体之间相关性的人为产物。已经提出了几种方法来解决这个问题，并在公开可用的软件包中实现。 环境因素往往与遗传变异相互作用，增加疾病的风险。识别这些相互作用，称为基因-环境（GxE）相互作用，现在是人类研究和模式生物研究的主要焦点。发现GxE相互作用可以深入了解疾病途径，了解环境因素对疾病的影响，更好的风险预测和个性化治疗。模式生物如小鼠是研究GxE相互作用的理想环境，因为环境暴露可以仔细控制。不幸的是，在关联研究中，相关性会导致虚假的关联，同样的原因，相关性也会导致虚假的基因与环境的相互作用。 在这项提案中，我们建议开发方法，纠正相关性的研究，寻找基因与环境的相互作用。我们项目的结果将是一套方法，即使研究中的个体是相关的，也可以持续检测基因与环境的相互作用。这些方法可以被许多参与研究的研究人员广泛使用，以发现基因与环境的相互作用。 我们将把我们开发的方法应用到明尼苏达州双胞胎和家庭研究中心（MCTFR）的数据中，以研究基因-环境相互作用如何影响药物滥用（SA）的发展，并应用到小鼠遗传研究中，研究影响高脂饮食反应和心力衰竭易感性的遗传因素。我们将通过公开的软件包和网络服务器资源向研究社区提供我们的方法。

DESCRIPTION (provided by applicant): Over the past few years genome-wide association studies (GWASs) have identified numerous genes associated with common human diseases. In these studies, genetic variation in thousands of individuals is collected and correlated with the disease status in these individuals. A challenging aspect of GWAS is that the collected individuals are related to each other by differing degrees. This can lead to spurious associations which are genes that appear to be associated with the disease, but in fact are an artifact of the relatedness between individuals. Several methods have been proposed to address this problem and are implemented in publicly available software packages. Environmental factors often interact with genetic variation to increase risk of disease. Identifying these interactions, referrd to as gene-by-environment (GxE) interactions, is now a major focus of research in both human studies and model organism studies. Discovering GxE interactions can provide insight into disease pathways, an understanding of the effect of environmental factors in disease, better risk prediction and personalized therapies. Model organisms such as mouse are ideal environments for studying GxE interactions because environmental exposures can be carefully controlled. Unfortunately, for the same reasons that relatedness can cause spurious associations in association studies, relatedness can cause spurious gene-by- environment interactions. In this proposal we propose to develop methodology that corrects for relatedness in studies that search for gene-by-environment interactions. The results of our project will be a set of methods that are can detect gene-by-environment interactions consistently even when the individuals in the study are related. These methods can then be widely used by many researchers involved in studies to discovery gene-by-environment interactions. We will apply our developed methods to the Minnesota Center for Twin and Family Research (MCTFR) data to investigate how gene-environment interplay influences the development of substance abuse (SA) and to mouse genetic studies investigating the genetic factors which influence response to high fat diet and susceptibility to heart failure. We will make implement our methods available to the research community through publicly available software packages and webserver resources.