Heterogeneous short-term plasticity reformats latent temporal information into an explicit population clock
In the authors' words
Short-term plasticity (STP) supports diverse synaptic-level computations, but how its organization across a circuit shapes emergent, network-level dynamics remains poorly understood. Here, we first characterize STP across sparse inputs to Kenyon cells in the Drosophila mushroom body, revealing an unstructured and near-random organization that could not be explained by either pre- or postsynaptic cell identity. Rather than representing biological noise, an experimentally constrained network model instead shows how this heterogeneous organization reformats temporal information embedded in presynaptic spike rate trajectories into a behaviorally relevant population clock. Although an ideal observer could recover the same degree of temporal information from most alternative models, only the STP-mediated transformation made that information accessible to a biologically realistic decoder. Together, our results identify heterogeneous STP as a distributed biophysical mechanism that collectively functions as a temporal basis set to support circuit computations across variable input statistics.
Appeared: Thursday, November 19. bioRxiv. Preprint, not yet peer-reviewed.