Speaker
Description
Building virtual cells and virtual tissues that capture how interacting cell populations evolve in space and time is a central goal of modern biology, yet spatial omics experiments can only provide sparse snapshots of this process. Here, we introduce CytoBridge, a computational method that reconstructs continuous virtual cell dynamics from discrete spatial transcriptomic data. Starting from a joint expression–space manifold, CytoBridge learns stochastic trajectories of cell states, positions, and population sizes through an unbalanced mean-field Schrödinger bridge formulation that explicitly incorporates cell–cell interaction as a dynamical driver. A time-varying interaction graph built from spatial proximity and ligand–receptor priors allows the transition velocity to be decomposed into an interaction program, learned from within-
system observables, and an intrinsic-context program that captures intrinsic regulatory processes. Applied to five datasets across three spatial transcriptomics platforms and spanning development, regeneration, and neurodegeneration, CytoBridge accurately generates tissue dynamics at unmeasurable time points, recovers known lineage trajectories and growth patterns, and reveals spatiotemporal ligand–receptor signalling axes not accessible from static analyses. CytoBridge provides a general method for turning heterogeneous spatiotemporal transcriptomic measurements into continuously evolving virtual tissues.
