Genetic and environmental interaction analyses and risk prediction modeling

项目总结（参见说明）：N2：项目3 卵巢癌是一种复杂的疾病，可能是多种基因和环境因素单独或联合导致其病因。虽然常见的遗传多态性和公认的生殖/生活方式风险因素的独立影响是病因之谜的一个重要组成部分，但直到系统地探索多位点效应（基因-基因相互作用）和生殖/生活方式因素与基因型之间的相互作用，这个谜才能完整。该项目的目标是对GxG和GxE效应进行系统评价，将最先进的分析程序应用于迄今为止建立的最全面的卵巢癌易感性数据集。特别是，我们计划在本研究的每个阶段将由项目1收集和维护的个人水平遗传数据与卵巢癌协会联盟（OCAC）维护的核心流行病学数据结合起来，以创建一个分析数据集，其中包含超过10，000例病例和10，000例对照，为遗传和环境相互作用效应提供信息。流行病学数据的汇编工作正在进行中。具体目标如下：1）对4个最大的组织学亚型中的每一个使用不可知和基于途径的方法来评估基因-基因相互作用：2）对14个流行病学因素，对4个最大的组织学亚型中的每一个使用不可知和基于途径的方法来评估基因-环境相互作用; 3）建立一个综合模型，用于评估发生卵巢癌的绝对风险，该模型包括家族史（BRCA 1/2状态）、其他公认的遗传和环境因素以及已发现的（项目1和3）和功能特征 （2）目前的研究。目标1和目标2共享一个共同的两级分析计划：由现有互补方法组成的屏幕，然后对屏幕上出现的显着相互作用效应进行基于模型的评估。该筛选的目的是评估和粗略优先考虑尽可能多和不同的一组多位点或多变量GxE效应，以进行更正式和计算要求更高的多项式响应、多变量调整分析。Aim 3风险模型将根据患者的卵巢癌风险是否由家族易感性（BRCA 1/2）或常见的流行病学和遗传风险因素驱动对患者进行分层。目标3的模型构建将侧重于指定后一个组件，并纳入本研究中确定的G、GxG和GxE效应。

PROJECT SUMMARY (See instructions): N2: PrOJect 3 Ovarian cancer is a complex disease and it is likely that multiple genes and environmental factors contribute, alone and in combination, to its etiology. While the independent effects of common genetic polymorphisms and of the accepted reproductive/lifestyle risk factors are an important part of the etiolgic puzzle, the puzzle will not be complete until the role of multi-locus effects (gene-gene interactions) and interactions between reproductive/lifestyle factors and genotype are systematically explored. The goal of this project is to implement a systematic evaluation of GxG and GxE effects that applies the most advanced analytic procedures available to the most comprehensive data set on ovarian cancer susceptibility built to date. In particular, we plan to join the individual-level genetic data at each phase of this study, assembled and maintained by Project 1, with the core epidemiolgic data maintained by the Ovarian Cancer Association Consortium (OCAC) to create an analysis data set with more than 10,000 cases and 10,000 controls informative for genetic and environmental interaction effects. Compilation of the epidemiologic data is underway. The specific aims are as follows: 1) To evaluate gene-gene interactions using both agnostic and pathway-based approaches for each of the four largest histological subtypes; 2) to evaluate gene-environment interactions, for 14 epidemiologic factors, using both agnostic and pathwaybased approaches for each of the four largest histological subtypes; 3) to create a comprehensive model for absolute risk of developing ovarian cancer that incorporates family history (BRCA1/2 status), other accepted genetic and environmental factors, and those discovered (Projects 1 and 3) and functionally characterized (Project 2) by the current study. Aims 1 and 2 share a common two-level analysis plan: a screen comprised of existing complementary methods followed by model-based evaluation of the notable interaction effects that emerge from the screen. The screen's purpose is to evaluate and coarsely prioritize as many and as diverse a set of multi-locus or multivariable GxE effects for more formal and computationally demanding multinomial response, multivariate adjusted analysis. The Aim 3 risk model will stratify patients according to whether their risk of ovarian cancer is driven by familial susceptibility (BRCA1/2) or common epidemiological and genetic risk factors. Model building for Aim 3 will focus on specifying the latter component and incorporating G, GxG and GxE effects identified in this study.