Research-lab-KUMS / working-memory-with-self-repairing

This repository is the MATLAB codes for the WM with SR model, published in iScience journal.

Working Memory (WM) with Self Repairing (SR) in a Spiking Neuron-Astrocyte Network (SNAN)

Some parts of the code are from: https://github.com/altergot/neuro-astro-network

The self-repairing capability and some modifications in the working memory model are added in this work.

This repository is the MATLAB codes for the WM with SR model, published in iScience journal.

Link to the paper: https://doi.org/10.1016/j.isci.2023.108241

Requirements:

• MATLAB 2018 or later.
• The minimum required amount of RAM is 32 GB

Settings

There are 4 experiments in the paper which require a uniques set of settings to get them.

Experiment 1 - Healthy network with SR signal

To get the results for this experiment set the following parameters in model_parameters.m file in this format:

params.SelfRepair = 1; % activates the SR modulation
params.simPattern = 1; % set the inputs as numbers 0 to 3
params.impairmode = 0; % keep the network in healthy condition

In order to get the results, run main.m file.

In order to plot the outputs, in Plots.m file, set the variables you want to plot to 1 and run Plots.m file.

Experiment 2 - Discrete pattern-specific damage mode

To get the results for this experiment set the following parameters in model_parameters.m file in this format:

params.SelfRepair = 1; % activates the SR modulation
params.simPattern = 1; % set the inputs as numbers 0 to 3
params.impairmode = 3; % damages the network with discrete lines damaging pattern

In order to get the results, run main.m file.

In order to plot the outputs, in Plots.m file, set the variables you want to plot to 1 and run Plots.m file.

Experiment 3 – Continuous pattern-specific damage mode

To get the results for this experiment set the following parameters in model_parameters.m file in this format:

params.SelfRepair = 1; % activates the SR
params.simPattern = 2; % set the inputs as two strips
params.impairmode = 3; % damages the network with dense core damaging pattern

In order to get the results, run main.m file.

In order to plot the outputs, in Plots.m file, set the variables you want to plot to 1 and run Plots.m file.

Experiment 4 – Analytic results for random damage mode

To get the results for this experiment set the following parameters in model_parameters.m file in this format:

params.SelfRepair = 1; % activates the SR
params.simPattern = 1; % set the inputs as numbers 0 to 3
params.impairmode = 2; % keep the network in healthy condition

In order to get the results, run analytic_metrics.m file. This simulation, depending on your PC configurations, takes around 16 hours to be simulated.

Since this simulation takes a long time to be done, we incorporated the saved output from our own simulations in dataa folder, in order to load those data and plot the outputs, you have to run analyticPlot.m file.

To get the code and data to run this experiment, contact the corresponding author "Mahmood Amiri".

Authors

• Pedram Naghieh - Implementation, Biological model constructing - PedRaMNG
• Abolfazl Delavar - Implementation, Biological model constructing - abolfazldelavar
• Mahmood Amiri - Project vision, Biological model constructing
• Herbert Peremans - Project vision

Cite

P. Naghieh, A. Delavar, M. Amiri, and H. Peremans, “Astrocyte’s self-repairing characteristics improve working memory in spiking neuronal networks,” iScience, p. 108241, Oct. 2023, doi: 10.1016/j.isci.2023.108241

License

This project is licensed under the MIT License - see the LICENSE.md file for details.