Cell monolayers as computing networks: predicting collective migration and pattern formation

Multicellular assemblies collectively exhibit physical intelligence through the development and maintenance of complex morphologies at the tissue scale without centralized control. During morphogenetic events, cell collectives must navigate the space of possible morphologies through the coordination of individual cell actions. We present a framework in which confluent cell collectives are modeled as networks of mechanically coupled computing agents. Each cell senses local network information (stress, substrate topography, etc.) and performs internal computations that direct decisions like migration and rearrangement. We explore the implementation of local physics-informed rules, such as biased cell rearrangements and directed active motion, and characterize the emergent phenomena. In particular, we investigate the formation of cell streams on anisotropic substrates, cell alignment under flow, and the migration of aligned nematic flocks, validating our predictions against experimental data.