Toward Diagnostics and Therapies of Molecular Subcategories of CAD

 描述(申请人提供)：冠状动脉疾病(CAD)是全球和美国的主要死亡原因。虽然这种疾病的遗传学本质上是复杂的，但由于过去5-10年的巨大研究投资，特别是在全基因组关联研究(Gwas)方面，人们已经实现了对CAD的更公正、数据驱动和现实的观点。作为这一成果的一部分，已确定了约160个共同的CAD/心肌梗死(MI)危险基因。现在一个重要的任务是了解这些基因座对CAD/MI产生风险的分子机制/途径，从而将最初的发现转化为新的治疗和诊断方法。然而，由于到目前为止识别的基因座只解释了约10%的CAD/MI风险变异，因此定义与GWA基因座平行运行的额外CAD途径也是至关重要的。近年来，考虑中间表型(DNA和疾病之间)的临床研究极大地加强了对GWA数据集中确定的风险基因的解释。此外，可以从中间分子表型识别的疾病网络为识别新的CAD途径和新的CAD疗法的靶点提供了基本的框架。在过去的6年里，我们进行了一项临床研究，考虑了冠心病患者的许多中间表型(STARnet研究)。在这项建议中，我们打算使用STARnet研究中新产生的DNA基因型和RNA序列数据来识别冠心病背后的动脉粥样硬化和代谢网络。然后，我们提出了一项新的CAD前瞻性研究(NGS预测研究)，主要目的是验证STARnet研究的结果。我们假设冠状动脉病变的范围和稳定性，因此可以通过定义关键的动脉粥样硬化基因网络的状态来准确地评估临床结果。反过来，活跃在肝脏、腹部脂肪和骨骼肌中的代谢网络影响动脉粥样硬化基因网络的状态。此外，从容易获得的组织(例如血液、皮下脂肪和血浆)中分离出来的分子数据可以用来识别可以预测冠心病引起的临床事件风险的生物标记物。为了检验这些假设，我们提出了以下具体目标。目的1：利用STARnet数据集和CARDIoGRAM荟萃分析GWA数据集，识别与CAD和/或CAD亚型因果连锁的调控贝叶斯基因网络。目的：通过对STARnet病例的DNA基因型、RNA序列和CAD血浆蛋白数据进行机器学习，找出预测冠心病临床事件的生物标志物(反映在句法评分中)。目的3：验证已确定的因果CAD eQTL/网络和生物标记物使用在山上进行的NGS预测研究。西奈医院，瑞典双胞胎研究和CAD细胞和动物模型。我们相信，建议的研究可以大大提高对CAD的分子理解，从而服务于更长期的目标，即对确诊为明确分子亚类的CAD患者进行预防性和个性化治疗。

 DESCRIPTION (provided by applicant): Coronary artery disease (CAD) is a leading cause of death worldwide and in the US. While the genetics of this disease are intrinsically complex, thanks to huge research investments during the last 5-10 years, particularly in genome-wide association studies (GWAS), a more unbiased, data-driven and realistic view of CAD has been achieved. As part of this achievement, ~160 common risk loci for CAD/myocardial infarction (MI) have been identified. An important task is now to understand the molecular mechanisms/pathways by which these loci exert risk for CAD/MI allowing to translating the initial findings into new therapies and diagnostics. However, since the loci identified thus far explain only ~10% of variation in CAD/MI risk, it is also essential to define additional CAD pathways operating in parallel with GWA loci. In recent years, clinical studies that consider intermediate phenotypes (between DNA and disease) have greatly enhanced interpretations of risk loci identified in GWA datasets. In addition, disease networks that can be identified from intermediate molecular phenotypes provide an essential framework to identify novel CAD pathways and targets for new CAD therapies. Over the last 6 years, we have performed a clinical study considering many intermediate phenotypes in CAD patients (the STARNET study). In this proposal we intend to use newly generated DNA genotype and RNA sequence data from the STARNET study to identify atherosclerosis and metabolic networks underlying CAD. We then propose a new prospective study of CAD (the NGS-PREDICT study) with the main purpose of validating findings from the STARNET study. We hypothesize that the extent and stability of coronary lesions, thus clinical outcomes can be accurately assessed by defining the status of key atherosclerosis gene networks. In turn, metabolic networks active in liver, abdominal fat, and skeletal muscle influence the status of the atherosclerosis gene networks. In addition, molecular data isolated from easily obtainable tissues (e.g., blood, subcutaneous fat and plasma) can be used to identify biomarkers that can predict risk for clinical events caused by CAD. To test these hypotheses, we propose the following specific aims. Aim 1: To identify regulatory Bayesian gene networks causally linked to CAD and/or CAD sub-phenotypes using the STARNET datasets and the CARDIoGRAM meta-analysis GWA datasets. Aim 2: Identify biomarkers predicting clinical events of CAD (reflected in SYNTAX score) by applying machine learning on DNA genotype, RNA sequence and CAD plasma protein data from easily obtainable tissues of the STARNET cases. Aim 3: To validate the identified causal CAD eQTLs/networks and the biomarkers using the NGS-PREDICT study performed at the Mt. Sinai Hospital, the Swedish Twin study and CAD cell and animal models. We believe the proposed studies can lead to a significantly better molecular understanding of CAD and thus, serve the more long-term goal of preventive and personalized therapies of CAD patients diagnosed in well-defined molecular subcategories.