Statistical Methods for Data Integration and Applications to Genome-wide Association Studies

摘要 大规模流行病学研究，包括生物库和全基因组关联研究 （GWAS），现在正迅速导致复杂疾病的新风险因素的识别。 现在有越来越多的机会来开发疾病风险的综合模型 结合遗传标记、其他生物标记、生活方式因素和社会人口统计学 指标然而，由于所有潜在风险的信息， 在一项足够大的研究中，往往得不到这些因素。相反，信息可能 可从不同的研究，其中每一个可能包括一些子集的期望 变量此外，由于个人层面数据的后勤和隐私问题， 摘要级信息，即，模型参数的估计，可以从一些 问题研究我们建议开发一系列新颖的统计方法， 整合不同的数据集，以解决遗传学等领域面临的现代问题 广泛地说，是观察性流行病学研究。在目标1中，我们将开发一个通用框架， 使用来自主研究的详细协变量数据构建逻辑回归模型， 纳入外部研究的汇总统计信息。我们将开发一系列 将此框架应用于GWAS，我们将使用协变量数据，包括高- 生物样本库中的生物标记物，并与外部汇总进行联合分析- 统计数据，为基因的效应修饰和介导提供有力的探索 协变量关联。在目标2中，我们将扩展目标1的拟议框架， 开发具有正则化参数估计的高维协变量的模型。我们 将开发所提出的方法的应用，用于精细定位和多基因风险评分 协变量条件分析。在目标3中，我们将进一步开发多个新型应用程序 数据整合框架的不同，以解释疾病结局的不同准确性/深度 不同研究的数据。我们将举例说明所提出的方法的应用风险 多种癌症（乳腺癌、黑色素瘤和肺癌）的建模，两种心脏代谢特征（2型 糖尿病和冠状动脉疾病）和精神疾病（重度抑郁症） 使用来自英国生物银行研究和乳腺癌协会的个体水平数据 联合会和外部GWAS汇总统计。我们将免费分发、开发和 分发用户友好的软件。

Abstract Large-scale epidemiologic studies, including biobanks and genome-wide association studies (GWAS), are now rapidly leading to the identification of novel risk factors for complex diseases. There is now increasing opportunity to develop comprehensive models for disease risk incorporating genetic markers, other biomarkers, life-style factors and sociodemographic indicators. There are, however, major challenges as information on all of the potential risk factors are often not available in a single adequately large study. Instead, information may be available from different studies, each of which may include some subsets of the desired variables. Further, because of logistical and privacy concerns with individual level data, only summary-level information, i.e., estimates of model parameters, may be available from some studies. We propose to develop a series of novel statistical methods that will allow data integration across disparate datasets to tackle modern problems faced in genetics and more broadly, observational epidemiologic studies. In Aim 1, we will develop a general framework for building logistic regression models using detail covariate data from a main study, while incorporating summary-statistics information from an external study. We will develop a series of applications of this framework to GWAS where we will use covariate data, including high- throughput biomarkers, from biobanks and perform combined analysis with external summary- statistics data for powerful exploration of effect modification and mediation of genetic associations by covariates. In Aim 2, we will extend the proposed framework of Aim 1 for developing models with high-dimensional covariates with regularized parameter estimates. We will develop application of the proposed method for fine-mapping and polygenic risk score analysis conditional on covariates. In Aim 3, we will further develop multiple novel applications of the data integration framework to account for different accuracy/depth of disease outcome data across different studies. We will illustrate application of the proposed methods for risk modeling of multiple cancers (breast, melanoma and lung), two cardiometabolic traits (type-2 diabetes and coronary artery disease) and a psychiatric disorder (major depression disorder) using individual level data from the UK Biobank study and Breast Cancer Association Consortium, and external GWAS summary-statistics. We will distribute develop and freely distribute user friendly software.