中央研究院 Logo

Mendelian Randomization (MR) has been widely used to infer causal relationships between exposures and outcomes in the genome-wide association study (GWAS) era. In recent years, substantial progress has been made in developing robust MR methodologies and expanding their applications. In this talk, I will present several methodological advances from my lab, including a bias-corrected multivariate MR framework and its extension incorporating transfer learning to strengthen causal inference in underpowered populations. I will also discuss our multivariable transcriptome-wide association study approach for identifying tissue-specific causal genes and pinpointing direct causal variants, as well as our work on detecting gene-environment interactions. All of these studies are built upon the MR framework.