Biostatistics and Data Science Facility Core

项目摘要 西奈山P30生命周期健康与环境中心（HEALS）的使命是 利用生命历程知情的健康模型加速团队科学研究。生物统计学和 数据科学设施核心（BDSFC）通过对复杂的暴露进行建模， 中心成员在一系列研究类型（基础，临床， 流行病学）、暴露和健康结果。服务包括研究设计、协变量选择、数据 分析以及开发方法时，需要解决我们中心的研究主题（混合物/ 生物学、临床环境研究和环境正义）。环境条件不是 平均分布，更高毒性暴露的风险不是随机的。空气质量差的比率较高， 水质、营养不良和接触有毒化学品的可能性沿着种族和 社会经济梯度通过拥抱生命历程的方法，我们的中心促进研究，解决 早期生活和后期生活环境都决定了发育健康轨迹。我们也 强调跨越空间（地理空间变异性）和时间的环境健康的复杂性 （纵向效应、生命阶段效应），同时将公共卫生（预防疾病）与医学联系起来 （诊断和治疗差异）。我们的方法基于"所有疾病"这一基本原则， 有环境基础”。虽然我们提供标准的数据分析服务（例如，线性模型，纵向 混合效应模型、把握度计算、研究分析计划）统计方法和研究设计 分析我们中心大部分工作中出现的复杂、高维数据所需的数据库仍然是 相对较新，需要先进的统计技术知识和能力，以策展和解释 复杂的生物学数据-由BDSFC的教员维护的专业知识。核心教师和研究人员 从事研究的问题和方法的挑战，从中心产生的动机 我们的中心核心的合作和创新。核心教职员工提供统计和数据科学 培训从事环境健康科学（EHS）相关项目的博士后研究员。提供 经由核心设施的这种服务允许P30中心构建和维护专用资源（例如， 计量和先进统计、数据科学和流行病学方法方面的专门知识， 环境混合物的评价）。因此，为了支持P30中心的总体目标，BDSFC 提出了以下具体目标：（1）确保中心项目的基础是健全的 生物统计学/数据科学原理，并使用最先进的方法设计和分析EHS数据;（2） 进行关键任务生物统计/数据科学方法研究，以进一步保证所有 研究和数据分析方法;以及（3）协助生物统计/数据科学原则的培训 和分析方法中心的研究人员，研究员和博士后学员。

Project Summary The mission of Mount Sinai's P30 Center on Health and Environment Across the LifeSpan (HEALS) is to accelerate team science–based research utilizing life course–informed models of health. The Biostatistics and Data Science Facility Core (BDSFC) plays a key role in that mission by modeling complex exposure and phenotype data generated by Center Members across a spectrum of study types (basic, clinical, epidemiologic), exposures and health outcomes. Services include study design, covariate selection, data analysis as well as developing methods when needed to address our Center's research themes (mixtures/ exposomics, clinical environmental research, and environmental justice). Environmental conditions are not distributed equally and the risk of higher toxic exposures is not random. Higher rates of poor air quality, poor water quality, poor nutrition and the probability of exposure to toxic chemicals tracks along racial and socioeconomic gradients. By embracing a life course approach, our Center promotes research that addresses both the early life and later life environments that determine developmental health trajectories. We also emphasize the complexity of environmental health which crosses space (geospatial variability) and time (longitudinal effects, life stage effects) while bridging public health (prevention of disease) with medicine (diagnosis and treatment variability). Our approach is based upon the fundamental principle that “all diseases have an environmental basis.” While we offer standard data analytic services (e.g., linear models, longitudinal mixed effects models, power calculations, study analysis planning) the statistical methods and study designs needed for analyzing the complex, high-dimensional data that arise in much of our Center's work are still relatively new and require knowledge of advanced statistical techniques and the ability to curate and interpret complex biologic data – expertise maintained by the faculty of the BDSFC. Core faculty and researchers engage in research motivated by questions and methodological challenges that arise from center collaborations and innovations by our Center Cores. Core faculty and staff provide statistical and data science training for postdoctoral fellows working on environmental health sciences (EHS) related projects. Providing such services via a core facility allows the P30 Center to build and maintain specialized resources (e.g., expertise in measuring and advanced statistical, data science and epidemiological methods related to evaluation of environmental mixtures). Thus, to support the overall goals of the P30 Center, the BDSFC proposes the following specific aims: (1) to ensure that Center projects are grounded in sound biostatistical/data science principles and use state-of-the-art methods for design and analysis of EHS data; (2) to conduct mission-critical biostatistical/data science methods research for further quality assurance of all research and data analysis methods; and (3) to assist in the training of biostatistical/data science principles and analysis methods to Center investigators, fellows and post-doctoral trainees.