Short-term synaptic depression can increase the rate of information transfer at a release site - Fig 1

(A) A binary asymmetric channel with a finite memory is used to model the release site under short-term depression. The synchronous release probability, pi, and the asynchronous release probability, qi, are determined based on the previous L release outcomes of the release site. (B) The algorithm for calculating the synchronous spike-evoked release probability, pi, given the last L release outcomes, (Yi−L, Yi−L+1, …, Yi−1). The asynchronous release probability, qi, can be derived similarly by substituting c, e and p with d, f and q respectively. We assume that the model parameters (α, c, d, e, f, p0, q0) are strictly greater than zero and less than one. The seed value of the algorithm, u0, is set to p0 for synchronous release and q0 for asynchronous release. (C) The performance of the algorithm is compared with the stochastic model of depression [22]. The stochastic model is based on the differential equation , where pr, τ, p0, u, and tr are the release probability, recovery time constant, default (maximum) release probability, depression coefficient and the release timing. At the bottom of the panel, the orange stem plot shows the timing of the release, tr, during an interval of 200 msec. The orange line demonstrates the release probability, pr, calculated from the differential equation. By assuming a time unit of Δ = 10 msec for the MRO model, the 200 msec interval corresponds to a memory of length L = 20. The memory content of the model is shown at the top of the panel. The blue line demonstrates the release probability calculated from the algorithm in (B). The final values of the release probability are indicated by filled circles. (D) Every arbitrary state of the release site can transit to two other states, depending on the release outcome. The transition probabilities are shown on the transition links.