报告人: 黄海燕(Department of Statistics, University of California, Berkeley)
时间:2024-12-22 11:00-12:00
地点:北京大学医学部新公卫楼220会议室
Abstract:
Precision medicine in oncology aims to identify relationships between specific cancer subtypes and the therapeutic compounds most effective against them. However, this task is complicated by the genetic heterogeneity and complexity of cancers, as well as the structural diversity of available drugs.
Here, we introduce ImpaCluster, an integrative deep multi-task prediction and biclustering framework that combines cancer omics, drug data, and drug response profiles to identify sensitive cancer-drug biclusters. ImpaCluster leverages shared genomic and molecular features of cancer cell lines and drugs to iteratively discover subsets of cell lines and drugs with heightened sensitivity. By utilizing local models incorporating latent embeddings of cell lines and drugs alongside drug response predictors, our approach reveals the molecular signatures that drive differential drug responses. Moreover, ImpaCluster’s ability to cluster unseen cell lines and compounds offers a rapid and scalable screening tool, distinguishing it from conventional biclustering methods. Validation through simulations and the Genomics of Drug Sensitivity in Cancer (GDSC) dataset highlights ImpaCluster’s utility in uncovering novel cancer-drug interactions and informing precision therapeutic strategies.
About the Speaker:
Haiyan Huang is a Professor and the Chair of the Department of Statistics at UC Berkeley. She served as the Director of the Center for Computational Biology at UC Berkeley from 2019 to 2022.
Prior to joining the faculty of the University of California, Berkeley in 2003, Haiyan Huang completed a two-year postdoctoral position in applied statistics and computational biology at Harvard University. She earned her Ph.D. in Applied Mathematics from the University of Southern California in 2001 and her B.S. in Mathematics from Peking University, China, in 1997.
As an applied statistician, her research focuses on the interface between statistics and data-rich scientific disciplines, such as biology. Over the past few decades, rapidly evolving biological technologies have generated enormously high-dimensional, complex, and noisy data, presenting increasingly pressing challenges to statistical and computational science. Her research group is devoted to addressing various modeling and analysis challenges posed by these data.

Your participation is warmly welcomed!

欢迎扫码关注北大统计科学中心公众号,了解更多讲座信息!