Biostatistics Department Seminar
Title: Making machines to learn biology from single-cell omics data
Abstract: Machine learning and other computational methods have been playing key roles in biological studies as essential tools for handling and analyzing genomics data. In recent years, we have been trying to explore possibility of designing machine learning models that are not tools for technical tasks, but can help mining the complex biology behind the data, especially from atlas-scale single-cell transcriptomics data. Our efforts include: (1) An attempt for developing an ab initio discovery learning method to form hypotheses from data without explicit human guidance for discovering cell lineages in early embryonic cell differentiation; (2) Designing and pre-training large AI cellular models (LCMs) as foundation models for a wide spectrum of single-cell analysis tasks such as data enhancement, drug-response prediction, perturbation effect prediction and for designated generation of pseudo scRNA-seq data; and (3) A mechanism-informed deep learning framework for prioritizing potential driver regulators for cell state transitions from scRNA-seq and scATAC-seq data. The efforts showed promising potential for making AI models to decipher the complex molecular and cellular system of cells, and also highlighted several major challenges.
Speakers
Xuegong Zhang is Professor of Pattern Recognition and Bioinformatics in the Department of Automation, Tsinghua University, and Adjunct Professor of the School of Life Sciences and School of Medicine. He is the current Vice President of the International Society for Computational Biology (ISCB) and a leading expert in AI, machine learning, large language models, and single-cell genomics.
Zoom Registration
If you would like to join via Zoom, please register here.
2024-2025 Monday Seminar Series
All seminars are held at 12:05 PM via Zoom and onsite. View all seminar information here.
