Statistical Methods for Medical Studies

描述（由申请人提供）：本更新申请建议开展统计方法研究计划，以解决三个生物医学研究领域出现的差距和障碍：项目1，“慢性病人群科学研究问题和障碍”，旨在开发与预防癌症和其他慢性病最相关的方法。这些包括用于分析多变量失败时间和纵向数据以及用于疾病风险归因的方法;用于使用生物标志物校正饮食和身体活动评估数据以及用于新生物标志物开发的方法;用于使用高维基因型数据来识别个体的优选治疗或干预的方法;以及用于生物网络开发和用于预防性干预开发的方法。项目2，“遗传流行病学方法”，主要侧重于在后基因组广泛关联研究时代更充分地了解遗传对疾病风险的贡献所需的方法。这些方法包括识别改变遗传效应的环境因素组合的方法;罕见变异关联研究的方法;遗传函数估计的方法;以及使用基因型数据促进环境因素关联研究的方法。项目3，“生物标志物在疾病诊断、预后、风险预测和早期检测中的应用”，建议开发用于预后生物标志物评估的新型研究设计，以通过使用标准化生物标志物值来改善ROC曲线的推断;并开发用于生物标志物评估的组序贯设计程序。总的来说，这些项目将以互动和协调的方式应用15名活跃的生物统计方法学家的才能，以解决对慢性病人群研究进展最重要的统计问题。

DESCRIPTION (provided by applicant): This renewal application proposes to carry out a Program of statistical methods research to address gaps and barriers arising in three biomedical research sectors: Project 1, 'Chronic Disease Population Science Research Issues and Strategies', aims to develop methods that are mostly pertinent to the prevention of cancer and other chronic diseases. These include methods for the analysis of multivariate failure time and longitudinal data, and for disease risks attribution; methods for correcting dietary and physical activity assessment data using biomarkers, and for new biomarker development; methods for using high dimensional genotype data to identify the preferred treatment or intervention for individuals; and methods for biological network development and for preventive intervention development. Project 2, 'Genetic Epidemiology Methods', focuses primarily on methods needed to more fully understand the genetic contribution to disease risk in the post-genome wide association study era. These include methods for identifying combinations of environmental factors that modify genetic effects; methods for rare variant association studies; methods for penetrance function estimation; and methods for using genotype data to facilitate environmental factor association studies. Project 3, 'Use of Biomarkers in Diagnosis, Prognosis, Risk Prediction and Early Detection of Disease', proposes to develop novel study designs for prognostic biomarker evaluation to improve inference on ROC curves through the use of standardized biomarker values; and to develop group sequential design procedures for biomarker evaluation. Collectively, these projects will apply the talents of 15 active biostatistical methodologists, in an interactive and coordinated manner, to address statistical issues that are among the most important for progress in chronic disease population research.