Estimating Short-term Synaptic Plasticity from Pre- and Postsynaptic Spiking

Fast information transmission in neural networks is heavily influenced by short-term synaptic plasticity (STP), and the type and timescale of STP varies by cell-type and brain region. Although STP has been widely characterized in vitrofrom recordings of postsynaptic potentials or currents, characterizing STP in in vivoin behaving animals is difficult due to the lack of large-scale intracellular recordings. Here, we use paired extracellular observations to estimate the short-term dynamics of synaptic transmission from spikes alone. We introduce an augmented generalized linear model (GLM) that includes a dynamic functional connection as well as several, non-synaptic factors that alter spike transmission probability. Our model captures the diverse short-term dynamics ofin vivospike transmission at identified synapses and accurately captures the effects of local pre- and postsynaptic spike patterns. We applied this model to large-scale multi-electrode recordings to describe stimulus-dependent shifts in spike transmission and cell-type specific differences in STP