Collaborative Research: Integrating Multi-Dimensional Omics Data for Quantifying Disease Heterogeneity

许多复杂的疾病，如癌症，表现出明显的跨患者异质性。为了更好地了解疾病生物学和选择最佳的治疗策略，正确地建立疾病异质性的模型是重要的。该项目将开发一个新的框架，通过有效集成来自多种类型的高度复杂的组学测量的信息来模拟疾病的异质性。拟议的分析框架和方法将产生重大而广泛的影响。这些方法的应用将有助于更准确地识别不同的患者群体及其组学特征，这将有助于疾病亚型的识别/定义、治疗选择和临床决策。关于皮肤癌和肺癌的数据将被分析，从而产生对基础科学研究人员和临床医生有价值的异质性模型。该项目还涉及耶鲁大学和爱荷华州大学研究生的教育和培训。高维组学数据已被证明对异质性分析非常有效。利用多维图谱的最新发展，在此基础上收集关于多种组学测量的数据，研究人员将系统地开发新的综合分析战略和方法。具体地说，将在新的PFR(惩罚融合回归)框架下开发三套方法。将进一步发展模型平均，以便于计算，并为拟议的方法提供更多的见解。将进行广泛而严格的方法、计算和理论研究。该项目将对高维统计和疾病异质性分析做出根本贡献。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Many complex diseases such as cancer demonstrate significant across-patient heterogeneity. For a better understanding of disease biology and optimally selecting treatment strategies, it is important to properly model disease heterogeneity. This project will develop a novel framework for modeling disease heterogeneity through the effective integration of information from multiple types of highly complex omics measurements. The proposed analysis framework and approaches will have significant broader impact. Applications of the methods will lead to more accurate identification of heterogeneous patient groups as well as their omics characteristics, which will facilitate the identification/definition of disease subtypes, treatment selection, and clinical decision-making. Data on skin and lung cancer will be analyzed leading to heterogeneity models that will be valuable to basic science researchers and clinicians. The project also involves education and training of graduate students at Yale University and the University of Iowa.High-dimensional omics data have been shown to be highly effective for heterogeneity analysis. Taking advantage of recent developments in multi-dimensional profiling under which data are collected on multiple types of omics measurements, the investigators will systematically develop novel integrated analysis strategies and approaches. Specifically, three sets of methods will be developed under the novel PFR (penalized fused regression) framework. Model averaging will be further developed to facilitate computation and provide additional insights into the proposed approaches. Extensive and rigorous methodological, computational, and theoretical investigations will be conducted. This project will make fundamental contributions to high-dimensional statistics and disease heterogeneity analysis.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.