Foundation Models for Genomics: From Epigenomic Prediction to Biological Reasoning
October 26 @ 1:00 pm – 2:00 pm
Statistical Bioinformatics Seminar
Speaker: Dr Wenpin Hou, Duke University
This is an online event held via Zoom.
Zoom link: https://uni-sydney.zoom.us/j/85114748391
Recent advances in foundation models and large language models (LLMs) are creating new opportunities for analyzing complex epigenomic data and accelerating biomedical discovery. However, important challenges remain in learning meaningful biological representations and ensuring that model predictions and reasoning are accurate, reliable, and biologically grounded. In this talk, I will present our research in two complementary directions. First, I will discuss the development of foundation models for genomic and epigenomic data, highlighting how deep learning can capture complex relationships between genomic context and DNA methylation to enable genome-wide epigenomic prediction without requiring direct methylation measurements. Second, I will discuss our efforts to apply and systematically evaluate LLMs for biological reasoning and computational analysis, including cell-type annotation, biological knowledge assessment, and biomedical image classification. I will highlight both the capabilities and limitations of current models, as well as approaches to improve their reliability through biological knowledge and ontology grounding.
Subscribe to our seminar mailing list
Find out more about the Statistical Bioinformatics seminar series
Dr Wenpin Hou
Dr Wenpin Hou is an Assistant Professor in the Department of Biostatistics and Bioinformatics at Duke University School of Medicine, affiliated with New York Genome Center. Her research lies at the intersection of computational biology, artificial intelligence, and statistical genomics, with a focus on developing computational methods to understand gene regulation using single-cell, spatial, and epigenomic data. More recently, her group has focused on foundation models and AI-driven approaches for genomics, aiming to build predictive models of cellular states and regulatory programs across diverse biological contexts. Prior to joining Duke, Dr Hou was an Assistant Professor in the Department of Biostatistics at Columbia University and an affiliated member of the Data Science Institute. She completed her postdoctoral training in Biostatistics and Computer Science at Johns Hopkins University and received her Ph.D. in Mathematics from The University of Hong Kong. Her work has been recognized through NIH Pathway to Independence Award (K99/R00) and Maximizing Investigators’ Research Award (MIRA) for Early Stage Investigators. She is also a recipient of the 2025 Mentor of the Year Award at Columbia University Irving Medical Center.