Mathematical Biology Seminar Xi Qiao, University of Utah, Wednesday, November 12, 2025 12:30 pm in LCB 222 Unraveling Causal Pathways: Bayesian Innovations for High-Dimensional Mediation in Omics Research Abstract: Modern biomedical studies generate vast, high-dimensional data from genomics and DNA methylation to the human microbiome, creating new opportunities to understand how biological systems respond to treatment and disease. Yet connecting these multi-omics profiles to clinical outcomes requires computational tools that move beyond association to uncover causal mechanisms. Causal mediation analysis provides one such framework by decomposing an exposure’s effect into direct and indirect pathways through intermediate biological processes. However, these omics mediators often comprise thousands of correlated features, rendering traditional low-dimensional models inadequate. This talk introduces a Bayesian framework for high-dimensional mediation analysis that integrates hierarchical modeling with sparsity-inducing priors to identify key mediators while accounting for the complexity and noise inherent in omics data. Applications in epigenetic and microbiome studies demonstrate how these methods reveal interpretable biological pathways and advance mechanism-driven discovery. Extensions toward flexible models and interdisciplinary applications will also be discussed.