Causal inference with Mendelian randomization
报告人： Yuehua Cui（Michigan State University）
时间：2026-04-09 15:10-17:00
地点：Room 217, Guanghua Building 2
Abstract:
Mendelian randomization (MR) leverages genetic variants as instrumental variables to estimate causal effects in observational settings. In this talk, I will discuss recent methodological advances in MR that strengthen its applicability to more complex data structures. First, I introduce MR-SPLIT, a framework for one-sample MR analysis that mitigates bias arising from IV selection and weak instruments by separating instrument selection and effect estimation under a data splitting framework. Second, I consider extensions of MR to longitudinal traits, where exposures and outcomes evolve over time, highlighting the challenges of time-varying confounding and dynamic causal effects. These approaches illustrate how instrumental variable methods can be adapted to improve robustness and interpretability in modern MR applications.
About the Speaker:
Yuehua Cui is a Professor in the Department of Statistics and Probability at MSU. His research focuses on statistical genomics and applied functional and longitudinal data analysis. He has published extensively in leading journals such as Nature Communications, Advanced Science, JRSS-B, AOAS and Biometrics. Dr. Cui is an elected Fellow of the American Statistical Association (ASA) and an elected member of the International Statistical Institute (ISI). He currently serves as an Academic Editor for PLOS Genetics and PLOS Computational Biology, and an Associate Editor for several journals in statistics and computational genomics including Statistics and Probability Letters and Statistical Applications in Genetics and Molecular Biology.
Your participation is warmly welcomed!
欢迎扫码关注北大统计科学中心公众号，了解更多讲座信息!
