You are here

Statistics Seminar: Qingyuan Zhao

Learning Instruments for Causality
January 28, 2019 - 3:30pm

101 Walter Library

Title: Learning Instruments for Causality

Abstract

Statisticians have long been at the forefront of making causal inference using observational data. In this talk I will share my experience with Mendelian randomization (MR), a causal inference method using genetic variants as instrumental variables. Due to the increasing availability of large-scale genome-wide association studies, MR is rapidly gaining popularity in epidemiology and human genetics. However, existing statistical methods still have several major limitations and lack theoretical grounding. I will illustrate how carefully designed data collection, statistical methods, and visual diagnostics can be instrumental in improving the quality of MR studies. Statistical theory is fruitfully applied to obtain highly efficient estimators with strong robustness to patterns of invalid instrumental variables occurring in real data. The new methods will be used to re-analyze several cardiometabolic diseases and risk factors, yielding new insights into the role of HDL particles (the "good" cholesterol) in coronary artery disease. To conclude the talk, I will sketch a “treasure map” of the pursuit of causality and describe how different statistical thinkings can help us to navigate this map. 
This talk is based on joint work with Jingshu Wang, Nancy Zhang, Dylan Small (University of Pennsylvania); Jack Bowden, Gibran Hemani, George Davey Smith (University of Bristol); Yang Chen (University of Michigan).

Bio
Qingyuan Zhao received his Ph.D. in Statistics from Stanford University in 2016 advised by Trevor Hastie and his B.S. in Mathematics from the Special Class for the Gifted Young (SCGY), University of Science and Technology of China (USTC) in 2011. He is currently a postdoctoral fellow in the Statistics Department of the Wharton School, University of Pennsylvania, mentored by Dylan Small and Sean Hennessy. His research interests include causal inference, high dimensional statistics, and applications in biomedical and social sciences.