Inferring Dynamic Cellular Trajectories and Underlying Cellular Regulatory Networks with Neural and Graph ODE Models - Part 1/2

Learn to infer dynamic cellular trajectories and underlying regulatory networks using advanced mathematical modeling techniques in this conference talk from Yale University's Smita Krishnaswamy at IPAM's Mathematics of Cancer workshop. Discover the PHATE method for embedding cellular data into low-dimensional state spaces while preserving data diffusion geometry, then explore how to infer cellular trajectories using MIOflow, a biologically principled neural ODE framework for manifold interpolating flows. Examine the mathematical foundations behind these computational approaches and understand their applications in cancer research, particularly in analyzing how cancer stem cells transform into aggressive metastatic states. Gain insights into cutting-edge methodologies that combine differential equations, neural networks, and geometric data analysis to model complex biological processes at the cellular level.