Cascades and convergence: dynamic signal flow in a synapse-level brain network

Connectomes—maps of synaptic connectivity—constrain how signals flow through the nervous system, shaping the transmission of sensory information to downstream targets involved in perception, decision-making, and action. Here, we use a simple, network-based spreading model to simulate sensory signal propagation across the adult Drosophila connectome. This approach allows us to trace modality-specific cascades and quantify their zones of overlap—neurons activated by multiple sensory pathways. Extending the classical spreading model, we introduce cooperative and competitive dynamics to simulate multisensory integration scenarios. Finally, we classify neurons based on their dynamical response profiles across all simulations, yielding a data-driven taxonomy grounded in both structure and dynamics. Our results highlight how abstract models can reveal organizing principles of neural computation and generate hypotheses for future experimental validation.