| Date and Time: |
Friday, October 23, 2026, 6:00 PM - 8:00 PM |
| Location: |
IRWD Community Room or Online
15500 Sand Canyon Ave., Irvine, CA 92618
|
RSVP via this Google form. A Zoom link will be sent out to all registrants.
Keynote Presentation
Generative Modeling of Cell-Type-Specific 3D Genome Organization
Diffusion models are a leading class of generative models behind many of today's image and video generators. They learn the underlying patterns of complex data and can generate new, realistic samples. In this talk, Dr. Wenxiu Ma will present a conditional diffusion model her research group is developing to deconvolve bulk 3D genomics data into cell-type-specific profiles. Most 3D genomics data are measured in bulk samples, which mix signals from many cell types or states. In contrast, her group's model uses other genomic data, such as single-cell gene expression, as the condition that guides the deconvolution toward each cell type, while preserving biologically meaningful features such as chromatin domains and loops.
About the Speaker
Wenxiu Ma is an Associate Professor in the Department of Statistics at the University of California, Riverside. She received her Ph.D. in Computer Science from Stanford University, followed by postdoctoral training in Genome Sciences at the University of Washington. Her research develops computational and statistical methods for biological big data. Her recent work focuses on 3D genome organization and deep learning methods for genomic data.