Emerging Statistical and Quantitative Issues in Genomic Research in Health Sciences

2018年数量基因组学（PQG）项目会议，题为“生物银行：研究设计和数据分析”，将于2018年11月1日至2日在哈佛医学院约瑟夫·b·马丁会议中心举行。这是三年来首次在定量基因组学（PQG）会议系列中获得该奖项的支持。这个长期的会议系列，向整个研究界开放，重点关注遗传和基因组科学中新兴的统计和计算挑战的及时跨学科讨论。每次会议的焦点都会发生变化，以反映科学前沿的发展。该系列成功的关键在于其跨学科性质，将定量和主题科学家聚集在一起，讨论在人类疾病的前沿遗传和基因组研究中出现的统计和定量问题。2018年会议主题的推动力来自全球范围内大规模生物银行的激增，这些生物银行由大量遗传和基因组数据、流行病学数据、电子病历、可穿戴设备和成像数据组成。生物银行的使用正在成为生物医学研究的一种必要的、潜在的革命性方法，它有能力改善各种疾病的预防、诊断和治疗，并促进个性化健康。本次会议旨在讨论生物样本库研究中的关键统计和定量挑战，包括满足广泛需求的生物样本库设计，提高表型准确性的策略，生物样本库之间的数据协调，以及分析生物样本库数据的新型统计和数据科学方法。会议向整个研究界开放，特别鼓励初级教师和研究人员、博士后、学生、妇女和代表性不足群体成员的参与。与会者将讨论和批评现有的定量方法，深入讨论新兴的统计和定量问题，并确定未来生物银行数据设计和分析研究的优先事项。所提出的研究将在科学期刊和网站的出版物中广泛传播。在2018年的PQG会议上，我们召集了一个跨学科的演讲者团队，他们在生物库的开发和研究设计方面经验丰富，包括统计学家、医学信息学家、遗传学家、流行病学家和临床科学家。目前有三个新兴的主题，包括人口生物银行的设计；生物库的表型和协调；以及生物银行数据分析。会议将围绕1)不同类型生物库的实例进行讨论，包括不同的招募策略，基因型/技术的选择，基因型资源和未来方向，设计策略：横断面，纵向，代表性概率抽样，增加多样性并扩展到美洲，非洲和亚洲；2)生物库之间的表型和协调问题，包括改善表型的新方法，生物库内部和生物库之间的协调，使用注册表数据丰富表型（事件时间），结合世界各地不同的生物库/注册表数据，偏差和缺失数据，电子健康记录（EHR）中的错误分类以及EHR数据库中的其他设计异质性问题；最后3)与数据分析方法相关的问题，包括如何使分析在计算上可扩展，相关样本和混合样本的分析，多表型和PheWAS分析，缺失数据，表型错误分类，处理某些疾病的少量病例，风险预测，选择适当的对照，跨生物库的综合分析，不同数据类型（基因型和表型）的综合分析，或转录组学和代谢组学数据)或在生物库中复制表型。欲了解更多信息，请参阅https://www.hsph.harvard.edu/2018-pqg-conference/This该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The 2018 Conference of the Program in Quantitative Genomics (PQG), entitled "Biobanks: Study Design and Data Analysis" will take place at the Joseph B. Martin Conference Center at the Harvard Medical School from November 1-2, 2018. This is the first of three years of the Program in Quantitative Genomics (PQG) Conference series to be supported by this award. This long-standing conference series, open to the whole research community, focuses on timely interdisciplinary discussions on emerging statistical and computational challenges in genetic and genomic science. The focus of each conference changes to reflect the evolution of scientific frontiers. Key to the success of the series is its interdisciplinary nature, bringing quantitative and subject-matter scientists together to discuss statistical and quantitative issues that arise in cutting-edge genetic and genomic research in human disease. The impetus for the 2018 conference theme comes from the proliferation of large-scale biobanks worldwide, composed of massive genetic and genomic data, epidemiological data, Electronic Medical Records, wearable devices and imaging data. The use of biobanks is becoming an essential and potentially revolutionizing approach to biomedical research, with the capacity to improve the prevention, diagnosis, and treatment of a wide range of illnesses, and to advance personalized health. This conference aims at discussing key statistical and quantitative challenges in biobank studies, including biobank design to meet a wide array of needs, strategies for improving phenotyping accuracy, data harmonization across biobanks, and novel statistical and data science methods for analysis of biobank data. The conference is open to the whole research community and particularly encourages participation of junior faculty and researchers, postdoctoral fellows, students, women and members of underrepresented groups. The participants will discuss and critique existing quantitative methods, discuss in-depth emerging statistical and quantitative issues, and identify priorities for future research in the design and analysis of biobank data. The research presented will be broadly disseminated in publications in scientific journals and websites. For the 2018 PQG conference we have assembled an interdisciplinary team of speakers experienced in biobank development and study design, including statisticians, medical informaticians, geneticists, epidemiologists, and clinical scientists. Three emerging topics of substantial current interest include the design of population Biobanks; phenotyping and harmonization across biobanks; and biobank data analysis. The conference discussion will revolve around 1) examples of different types of biobanks including different examples of recruitment strategies, selection of genotype / technologies, genotype resources and future directions, design strategies: cross-section, longitudinal, representative probability sampling, and increasing diversity and expanding to the Americas, Africa and Asia; 2) issues of phenotyping and harmonization across biobanks including new methods for improving phenotyping, harmonization within and across biobanks, enriching phenotyping using registry data (time to event), combining different biobank/registry data across the world, bias and missing data, misclassification in electronic health records (EHR) and other design heterogeneity issues in EHR databases; and finally 3) issues related to data analysis methods, including how to make analysis computationally scalable, analysis of related samples and admixture samples, multiple phenotypes and PheWAS analysis, missing data, misclassification of phenotypes, handling small number of cases for some diseases, risk prediction, selection of appropriate controls, integrative analysis across biobanks, and integrative analysis of different data types (genotype and phenotype, or transcriptomic and metabolomic data) or replicating phenotypes across biobanks. For more information, see https://www.hsph.harvard.edu/2018-pqg-conference/This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.