Polygenic Risk Score Diversity Consortium Coordinating Center

项目摘要 多基因风险评分（PRS）是从流行病学研究中的大规模基因组数据分析中开发的 作为一种精确医学工具，它有望帮助识别疾病风险较高的个体。但 目前，PRS研究和临床翻译面临的一个关键挑战是， 使用主要来自欧洲血统的个人的数据，导致预测性能较差， 非欧洲人群，反映了基因组和生物医学研究缺乏多样性， 大致上NHGRI正在建立一个新的多基因风险评分多样性联盟，以应对这一挑战， (1)利用现有的不同队列，改善对不同非欧洲人群的PRS预测 以及（2）优化大规模基因组和表型数据集的整合 支持合作分析、报告，以及为更广泛的科学研究创造方法和资源， 社区作为协调中心，我们将通过实现以下四个主要目标来支持联盟的目标：（1） 在联盟研究中心和附属成员之间协调基因型和表型数据， 包括变异和样本水平的质量控制、基因型插补和标准表型的使用 本体论(2)组织跨联盟分析和协作方法开发， 标准化评估减贫战略方法，以确定共识方法，发展地方血统 知情的PRS方法，以及将ELSI考虑因素整合到分析最佳实践中。(3)便于数据 通过NHGRI等社区资源和知识库在联盟内外共享 Genomic Data Science Analysis，Visualization，and Informatics Lab-space（AnVIL），NCBI数据库 基因型和表型（dbGaP），和多基因评分目录。(4)协调项目后勤和 外联，包括通过联合会网站、面对面会议、辅导和培训，以及定期 精准医疗合作伙伴计划。我们提出了创新的方法来实现这些目标 四个主要目标，包括利用新兴的基于云的平台进行协调、分析和共享 大规模基因组数据;研究可扩展的数据科学方法来协调表型; 领导一次全联合会范围的"烘烤"，以确定关于减贫战略制定和评价的协商一致办法; 并将伦理、法律的和社会影响纳入分析最佳实践。而且我们的 申请依赖于13年的成功经验和专业知识，作为五个协调中心， 大规模的基因和生物医学项目。通过科学和行政领导， 联合会，我们将帮助实现公共卫生利益的发展，预测和帮助预防或 减轻遗传多样性人群中的一系列疾病。

PROJECT SUMMARY Polygenic Risk Scores (PRS) developed from analysis of large-scale genomic data in epidemiological studies hold promise as a precision medicine tool that will help identify individuals at higher disease risk. However, a critical challenge now facing PRS research and clinical translation is that most PRS have been developed using data from predominantly European-descent individuals, resulting in poorer predictive performance in non-European populations and reflecting the lack of diversity in genomic and biomedical research more broadly. NHGRI is establishing a new Polygenic Risk Score Diversity Consortium to address this challenge by (1) leveraging existing diverse cohorts to improve PRS prediction across diverse, non-European populations and for a range of conditions and (2) optimizing the integration of large-scale genomic and phenotype datasets in support of collaborative analysis, reporting, and creation of methods and resources for the broader scientific community. As Coordinating Center, we will support the Consortium's goals by achieving four main aims: (1) leading harmonization of genotype and phenotype data across Consortium Study Sites and Affiliate Members, including variant and sample level quality control, genotype imputation, and use of standard phenotype ontologies. (2) Organizing cross-Consortium analysis and collaborative methods development, featuring standardized evaluation of PRS methods to identify consensus approaches, development of a local-ancestry informed PRS method, and integration of ELSI considerations into analytic best-practices. (3) Facilitating data sharing within and beyond the Consortium through community resources and repositories such as the NHGRI Genomic Data Science Analysis, Visualization, and Informatics Lab-space (AnVIL), NCBI database of Genotypes and Phenotypes (dbGaP), and the Polygenic Score Catalog. (4) Coordinating program logistics and outreach including through a Consortium website, in-person meetings, mentorship and training, and regular engagement with precision medicine partner programs. We propose innovative approaches to achieve these four main aims, including leveraging emerging cloud-based platforms for harmonization, analysis, and sharing of large-scale genomic data; investigating scalable data science approaches to phenotype harmonization; leading a Consortium-wide “bake-off” to establish consensus approaches to PRS generation and evaluation; and incorporating Ethical, Legal, and Social Implications into analytic best practices. Furthermore, our application rests on 13 successful years of experience and expertise serving as Coordinating Centers for five large-scale genetic and biomedical projects. Through scientific and administrative leadership of this Consortium, we will help realize the public health benefit of developing PRS that predict and help prevent or mitigate a range of diseases in genetically diverse populations.