报告名称：Boolean Landscapes and Hidden Biological Dynamics

时间：2026年06月22日 （周一）16:00 - 17:00

地点：集贤楼524

主讲人：曾辰教授

主办单位：物理与光电工程学院

主讲人简介：曾辰，美国乔治华盛斯顿大学(The George Washington University)物理系教授。目前致力于人工智能和计算生物物理方面的研究：从基于结构的药物设计到生物复杂性的大数据科学。曾辰教授的研究工作长期获得美国国家科学基金会和美国国立卫生研究院的资助. 在国际顶级刊物如自然免疫学(Nature Immunology), 美国科学院院报(PNAS), 物理评论快报(PRL), 德国应用化学(Angew Chem Int Ed Engl.)和生物信息学(Bioinformatics)等国际期刊上发表多篇学术论文. 曾辰教授于2001年获得美国国家自然科学基金会杰出青年教授奖。

活动简介：Modern biology often gives us snapshots of a system, but not the full movie of how the system changes over time. For example, cryo-EM can show the major shapes of a protein machine and how frequently each shape appears, but it usually cannot tell us the order in which those shapes change or whether there are hidden flows between them. Single-cell and spatial omics face a similar challenge: they can map different cell states and possible paths between them, but the actual regulatory rules driving those changes are often unknown.In this seminar, I will discuss our preliminary work on using Boolean network models and mean-field approximations to bridge this gap. The basic idea is to start from partial information about the observed steady-state landscape, build a family of simple network models consistent with that information, and then study the possible dynamics these models imply. This allows us to infer candidate transition pathways, probability landscapes, nonequilibrium fluxes, and Boolean-state patterns that may explain how the observed landscape is generated.