Pynn demo files to simulate networks of RSFS cells

Those PyNN demo files simulate a network of regularspiking (RS)
excitatory neurons and fastspiking (FS) inhibitory neurons. We study
here the network at the level of single cells, then his spontaneous
activity and finally its response to timevarying external input, as
described in the following paper:
Zerlaut Y, Chemla S, Chavane F, Destexhe A (2018) Modeling mesoscopic
cortical dynamics using a meanfield model of conductancebased
networks of adaptive exponential integrateandfire neurons. J Comput
Neurosci 44:4561
These files were contributed By Amelie Soler (Destexhe lab). Many
example output images are provided in the subfolders.
This paper presents a RSFS meanfield model of networks of Adaptive
Exponential (AdEx) integrateandfire neurons, with conductancebased
synaptic interactions. It uses a Master Equation formalism, together
with a semianalytic approach to the transfer function of AdEx neurons
to describe the average dynamics of the coupled populations. It
compares the predictions of this meanfield model to simulated
networks of RSFS cells, first at the level of the spontaneous
activity of the network. Second, it investigates the response of the
network to timevarying external input. Finally, to model VSDi
signals, a onedimensional ring model made of interconnected RSFS
meanfield units is used.
The simulations shown here are reproductions of Figure 2 (simulation
1), Figure 3a,b (simulation 2) and Figure 5a,b (simulation 3) of the
paper.
SImulation 1: It shows the response of the single cell models (RS in
green and FS in red) to an external current step of 200 pA lasting 300
ms. The parameters of the cells are the same as the ones presented on
Table 1 in the article. The membrane potentials are plotted with the
spikes (represented with dots) on the same graph.
Simulation 2: This script shows the spontaneous activity of the
network. The network is made of 8000 excitatory neurons and 2000
inhibitory neurons. Those two populations of neurons are randomly
connected (internally and mutually) with a connectivity probability of
5%. Plus a feedforward input (a ramp of 4HZ) on the inhibitory
population and on the excitatory population is added. We simulate the
network for 1000ms. The spiking activity and firing rate of the
network is plotted (green: excitatory neurons, red: inhibitory
neurons).
Simulation 3: This last script shows the response of the network to
timevarying external input. The duration of simulation and the
construction of the network is the same. Only here, the feedforward
input on the excitatory population is varying between 4Hz and 6H
(following theoretically a gaussian in the paper). During 100ms, it’s
the first part of the gaussian; here we simulate the gaussian used by
a ramp. Then for 200ms, it’s the second part of the gaussian, we
simulate it with a descending ramp. Identically, the spiking activity
and firing rate of the network is plotted (green: excitatory neurons,
red: inhibitory neurons).
If you use this for your research, please cite the above paper.
Amélie Soler
CNRS, Gif sur Yvette, France
http://cns.iaf.cnrsgif.fr
