Integrative statistical models for pathway analysis of GWAS data

描述(由申请人提供)：全基因组关联研究(GWAS)已成为识别复杂疾病背后的遗传变异的流行方法。然而，到目前为止发现的变异，无论是单独的还是结合在一起，只占疾病风险的遗传成分的一小部分。一个可能的原因是，复杂的疾病可能是由系统层面的变化引起的，例如在生物网络或途径中，个别基因变异对疾病风险只有微弱的边际影响。在这项提案中，我们将结合统计学、生物信息学和遗传学来开发旨在了解复杂疾病背后的遗传结构的综合方法。我们将检查最近发展的高密度基因分型阵列检测到的拷贝数变异(CNV)和单核苷酸多态(SNPs)，然后我们将结合先前的生物学知识来正式测试疾病与同一途径中的一组常见和罕见的遗传变异(CNV和SNPs)的联合作用，或者更广泛地说，基因集。我们将应用我们的新方法来分析两个精神分裂症(GAIN和非GAIN)和一个双相情感障碍(GAIN)GWAS数据集，所有这些数据集都是使用Affymetrix 6.0芯片生成的。我们的具体目标如下：(1)通过利用不同类型的遗传变异：常见和罕见的、CNV和SNPs的信息，开发新的统计方法来识别GWAs中具有丰富关联信号的基因和途径(或基因集)。我们将使用随机基因效应以层次化的方式对路径中的所有基因、SNPs和CNV进行建模，这提供了在同一路径中的基因之间借用信息的能力。(2)将所提出的模型应用于两种复杂的疾病(精神分裂症和双相情感障碍)，并开发一个用户友好的软件包来实现所提出的方法。《目标1》的成功完成将为我们目前和未来的全球人口基金研究提供重要的统计工具。Aim 2的成功完成将显著提高我们对精神分裂症和双相情感障碍背后的遗传结构，包括它们共同的遗传成分的理解，并将导致更有效的精神障碍治疗策略。

DESCRIPTION (provided by applicant): Genome-Wide Association Studies (GWAS) have become a popular approach for identifying genetic variants underlying complex diseases. However, the variants identified so far, individually or in combination, account for only a small proportion of the inherited component of disease risk. One possible reason is that complex diseases are likely to be caused by changes at the systems level, such as in a biological network or pathway, in which individual genetic variants only have weak marginal effects on disease risk. In this proposal, we will combine statistics, bioinformatics, and genetics to develop integrative approaches aimed at understanding the genetic architecture underlying complex diseases. We will examine copy number variants (CNVs) and single nucleotide polymorphisms (SNPs) detected by the recently developed high density genotyping arrays, and then we will integrate prior biological knowledge to formally test disease association with joint effects of groups of common and rare genetic variants (CNVs and SNPs) in the same pathway, or more broadly, gene set. We will apply our novel methods to analyze two schizophrenia (GAIN and nonGAIN) and one bipolar disorder (GAIN) GWAS datasets, all of which were generated using Affymetrix 6.0 chips. Our Specific Aims are as follows: (1) Develop novel statistical method to identify genes and pathways (or gene sets) with enriched association signals in GWAS by leveraging information from different types of genetic variants: common and rare, CNVs and SNPs. We will model all the genes, SNPs, and CNVs within a pathway in a hierarchical fashion using random gene effects, which provide the ability to borrow information across genes in the same pathway. (2) Apply the proposed model to two complex diseases (schizophrenia and bipolar disorder) and develop a user friendly software package that implements the proposed methodology. The successful completion of Aim 1 will provide us with critical statistical tools for current and future GWA studies. The successful completion of Aim 2 will significantly enhance our understandings of the genetic architecture underlying schizophrenia and bipolar disorder, including their common genetic components, and will lead to more effective treatment strategies for mental disorders.