The Neural Simulation Tool - NEST

This is a fork of NEST (version 3.1) in which the short-term-plasticity synapse named tsodyks3_synapse has been included. This version has been used in the context of the publication:

Tiddia, G., Golosio, B., Fanti, V., & Paolucci, P. S. (2022). Simulations of working memory spiking networks driven by short-term plasticity. Frontiers in Integrative Neuroscience, 16, 972055. https://doi.org/10.3389/fnint.2022.972055

You can find the GitHub repository of the Working Memory spiking model here.

The original NEST README can be found here.

Intalling this version

The installation has been tested on Ubuntu 21.10 and 22.04 by following these installation instructions. The link refers to the advanced Ubuntu/Debian installation instructions of the NEST Documentation of the version 3.1 of NEST, from which this version is derived.

Currently, this version can not be installed using PPA or conda.

Documentation

The Documentation to refer to is the NEST Documentation v3.1.

In the following is documented the new synapse model implemented in this version.

The tsodyks3_synapse model

The tsodyks3_synapse model comes from a modification of the tsodyks2_synapse model, already existing in NEST. The documentation for the tsodyks2_synapse model can be found here.

The difference between these two models stems from the order of update of the STP variables u and x when a spike is emitted. These variables modulates the synaptic efficacy of a synapse following the equation $J^{(\text{mod})}=J^{(\text{abs})}ux$, with $J^{(\text{abs})}$ being the absolute synaptic efficacy.

Given a spike emitted at a time $t_s$, in tsodyks2_synapse the modulation is obtained by multiplying the absolute synaptic efficacy with the values of $u$ and $x$ before the spike emission, i.e., $J^{(\text{mod})}(t_s)=J^{(\text{abs})}u(t_s^{-})x(t_s^{-})$, where $t_s^{-}$ refers to the time immediately before the spike emission.

Given the same spike emitted at a time $t_s$, in tsodyks3_synapse the modulation is obtained by $J^{(\text{mod})}(t_s)=J^{(\text{abs})}u(t_s^{+})x(t_s^{-})$, where $t_s^{-}$ is the time immediately before the spike emission and $t_s^{+}$ is the time immediately after the spike emission.

More details on the implementation of the model and on the motivation of this changes are reported in the publication:

Tiddia, G., Golosio, B., Fanti, V., & Paolucci, P. S. (2022). Simulations of working memory spiking networks driven by short-term plasticity. Frontiers in Integrative Neuroscience, 16, 972055. https://doi.org/10.3389/fnint.2022.972055

License

NEST is open source software and is licensed under the GNU General Public License v2 or later.