V1Novelty
A spiking neural network model to investigate cortical novelty responses
This project is concerned with the underlying mechanisms that give rise to novelty responses in the primary visual cortex as reported in Homann et al., BioRxiv 2017.
We set up a spiking neural network model of the mouse primary visual cortex (80% excitatory, 20% inhibitory) using various plasticity mechanisms. Some of the simulation code follows code provided by Litwin-Kumar and Doiron [1].
The code is written in Julia (https://julialang.org/, https://github.com/JuliaLang/julia).
Required packages:
- PyPlot
- HDF5
- Distributions
- Dates
- LinearAlgebra
- Random
Packages can be installed via using Pkg Pkg.add("PackageName").
The project is work in progress and will be updated continuously.
Project strucutre
main- contains initialisation files that start the simulation (to be run on cm) and evaluation (to be run locally - no plotting on cm possible)simulation- contains the main simulation and simulation helper filesevaluation- contains the main evaluation (weight, spiketime) and evaluation helper files
At this point plotting is still included directly in the evaluation files.
The data is stored on the server using the hdf5 file format.
To avoid many small files on the server the analysis files are stored locally.
[1] A. Litwin-Kumar & B. Doiron. Formation and maintenance of neuronal assemblies through synaptic plasticity. Nature Communications (2014).
Copyright notice:
http://lk.zuckermaninstitute.columbia.edu/
litwin-kumar_doiron_formation_2014
Copyright (C) 2014 Ashok Litwin-Kumar