Statistical Methods for Analysis and Integration in Genomic Studies of Disease

描述(由申请人提供)：基于基因组的疾病研究现在涉及到从大量患者中收集的高度多样化的数据类型。科学家面临的一个主要挑战是 如何最好地结合数据，提取重要特征，并全面描述它们影响个人病程和/或治疗反应可能性的方式。该项目旨在开发统计方法，以解决在基于基因组的疾病研究中出现的重要问题。特别是，我们提出了一些方法来提高从全基因组基因表达研究中获得的结果的能力和准确性。我们还提出了跨平台和多尺度集成数据的统计方法。这些综合方法能够实现与识别和量化随生物条件(例如，健康与疾病)变化的特征组相关的强大推理，并且它们还允许识别影响患者的病程和/或治疗反应的重要特征集合。该项目的成功完成将有助于确保从强大的基于基因组的技术中获得最大的效用，这些技术现在经常用于努力获得关于疾病表现、进展和维持的基因组机制的洞察和信息。 公共卫生相关性：发展统计上可靠的方法来解决复杂特征的基因组基础，对于个体化药物和改善公共健康至关重要。理想情况下，对患病个体的高通量遗传、基因组和表型测量将迅速导致识别其疾病背后的显著特征，以及这些特征如何影响疾病过程的规范。在实现这一理想之前，必须克服生物统计学中的许多挑战。这项提案解决了其中一些关键挑战。

DESCRIPTION (provided by applicant): Genomic based studies of disease now involve highly diverse types of data collected on large groups of patients. A major challenge facing scientists is how best to combine the data, extract important features, and comprehensively characterize the ways in which they affect an individual's disease course and/or likelihood of response to treatment. This project aims to develop statistical methods to address important problems that arise in genomic based studies of disease. In particular, we propose methods to improve the power and accuracy of results obtained from genome-wide studies of gene expression. We also propose statistical methods that integrate data across multiple platforms and scales. These integrative methods enable powerful inference related to identifying and quantifying groups of features that change across biological conditions (e.g. healthy vs. disease), and they also allow for the identification of important collections of features that affect a patient's disease course and/or treatment response. Successful completion of the project will help to ensure that maximal utility is gained from the powerful genomic-based technologies that are now routinely used in efforts to gain insights into and information about the genomic mechanisms underlying disease manifestation, progression, and maintenance. PUBLIC HEALTH RELEVANCE: The development of statistically sound approaches to resolve the genomic basis of complex traits is vital to individualizing medicine and improving public health. Ideally, high-throughput genetic, genomic, and pheno- typic measurements on diseased individuals would lead quickly to the identification of the salient features underlying their disease, along with a specification about how these features affect disease course. Many challenges in biostatistics must be overcome before this ideal is achieved. This proposal addresses some of those critical challenges.