报告题目：Integrating statistical evidence from big biomedical data
报 告 人：Yong Chen, Ph.D.
Associate Professor of Biostatistics
Department of Biostatistics, Epidemiology and Informatics
University of Pennsylvania
邀 请 人：李博安教授 纪志梁教授
联 系 人：彭永莹
报告摘要：With the increasing availability of medical and genomic data, it is critically important to effectively combine evidence, signals, or information from multiple data sources to enable reproducible scientific discovery. In this talk, I will share several stories from my experience in integrating big biomedical data. First, in pharmacovigilance it is critical to combine data from different sources for identifying safety signals for rare adverse events. I will talk about strategies in “pooling” data from large electronic health records (EHR) databases in different hospitals where the hospitals are not allowed to share individual patient level data due to privacy reasons. Secondly, I will share our recent work on integrating EHR data and genetic data from Penn Medicine Biobank for studying complex genetic architectures and their impacts on multiple phenotypes. Lastly, I will discuss some key challenges that have negatively impacted the reproducibility of research findings using EHR data, and introduce the current efforts in addressing these challenges. Real motivating examples will be introduced and discussed throughout the talk.