Data Management and Statistics Core

项目摘要 青少年节律、奖励和睡眠中心（CARRS）的核心假设是， 发育作用于潜在的睡眠和昼夜特征以改变稳态睡眠驱动，昼夜相位， 和昼夜节律的调整，这反过来又影响对物质使用风险至关重要的皮质边缘功能（例如，奖励 认知控制）。我们进一步假设，睡眠和昼夜节律的具体操纵， 青少年对奖赏反应性和认知控制有正向或负向的影响。这些 操作将为我们的模型提供实验支持，并为新的临床干预措施提供概念证明 减少药物使用和滥用的风险。核心C：数据管理和统计将支持5 通过管理数据（例如，制定方案、表格和数据库，并确保数据 质量和安全性）和执行统计分析（例如，初步、主要、次要和探索性 分析）。这样，Core C将保证数据的高质量、透明、一致的标准 管理和跨项目的统计分析，并将最大限度地提高CARRS内的严谨性和可重复性。 核心C还将开发和适应分析方法，充分利用翻译和高， 在CARRS中的5个项目中捕获的维度数据。重点领域将包括： 学习（例如，随机森林）用于预测人类和啮齿动物的高维数据; 跨项目和物种整合研究结果的方法（基于项目1-5的数据），以及定量 用于整合RNA-seq和质谱数据的“多组学”方法。最后，我们将教育研究人员 在CARRS和研究界中，了解与我们相关的现有和前沿统计方法 research.主题将包括严谨性和可重复性，睡眠和昼夜数据分析，RNA分析， 测序和蛋白质组学数据，以及核心C中开发和应用的创新统计方法。

PROJECT SUMMARY The central hypothesis for the Center for Adolescent Rhythms, Reward, and Sleep (CARRS) is that adolescent development acts on underlying sleep and circadian traits to modify homeostatic sleep drive, circadian phase, and circadian alignment, which in turn impact cortico-limbic functions critical to substance use risk (e.g., reward and cognitive control). We further hypothesize that specific manipulations of sleep and circadian rhythms during adolescence will affect reward responsivity and cognitive control in either positive or negative directions. These manipulations will provide experimental support our model, and proof of concept for novel clinical interventions to reduce the risk of substance use and abuse. Core C: Data Management and Statistics will support the 5 projects in CARRS by managing data (e.g., developing protocols, forms, and databases and assuring data quality and security) and performing statistical analyses (e.g., preliminary, primary, secondary, and exploratory analyses). In this way, Core C will guarantee high-quality, transparent, and consistent standards for data management and statistical analyses across projects and will maximize rigor and reproducibility within CARRS. Core C will also develop and adapt analytic methods that take full advantage of the translational and high- dimensional data captured across the 5 projects within CARRS. Areas of focus will include: use of supervised learning (e.g., random forests) for prediction with high-dimensional data within humans and, separately, rodents; methods for integrating findings across projects and species (based on data from Projects 1-5), and quantitative “multi-omic” methods for integrating RNA-seq and mass spectrometry data. Finally, we will educate researchers within CARRS and in the research community on existing and cutting-edge statistical methods relevant for our research. Topics will include rigor and reproducibility, analysis of sleep and circadian data, analysis of RNA sequencing and proteomic data, and the innovative statistical methods developed and applied within Core C.