ca1: a NEURON network model
ca1 is a large-scale, detailed, biologically constrained NEURON network model of the rodent hippocampal CA1 area.
- 1 Main features
- 2 Documentation
- 3 Installation
- 4 Usage
- 5 Environment variables
- 6 Outputs
- 7 Customizing
- 8 Meta
- Linux, Mac OS X and Windows support
Full documentation is available here.
On macOS, ca1 code can be installed via Homebrew (recommended):
$ brew install ...
A MacPorts port is also available:
$ port install ...
# Debian, Ubuntu, etc. using Docker $ apt-get install ....
# FDebian, Ubuntu, etc. manual install $ yum install ...
A universal installation method (that works on Windows, Mac OS X, Linux, …, and always provides the latest version) is to use `pip`_:
# Make sure we have an up-to-date version of pip and setuptools: $ pip install --upgrade pip setuptools $ pip install --upgrade ...
pip installation fails for some reason, you can try
easy_install ... as a fallback.)
$ nrngui main.hoc # Runs a network at scale 1:10000 for 100 ms to ensure all works # OR $ nrngui -c "SimDuration=600" main.hoc -c "quit()" # Runs a network at scale 1:10000 for 600 ms, long enough to see theta emerge # OR $ nrngui -nopython ... # if there is no python installed # OR $ nrniv ... # if using without a GUI popup (depends on your NEURON installation; sometimes nrngui must be invoked to ensure all necessary libraries are loaded when NEURON launches)
$ nrngui [-c parameters] [main.hoc file] -c "quit()" # adding the quit() command after ensures the code stops immediately after an error
Scaled down model to run directly on a personal computer:
$ nrngui main.hoc # Runs a network at scale 1:10000 for 100 ms to ensure all works # OR $ nrngui -c "Scale=1000" -c "SimDuration=600" main.hoc -c "quit()" # Runs a network at scale 1:1000 (will take a long time) for 600 ms, long enough to see theta emerge # OR $ nrngui -nopython -c "SimDuration=1000" -c "ConnData=446" main.hoc -c "quit()" # try a different connectivity configuration and a longer simulation with the default Scale of 1:10000, use the nopython flag if NEURON has errors due to not finding the site module, etc. # OR $ nrngui -c "Scale=1" main.hoc -c "quit()" # run a full scale network - this is not feasible for the ca1 network on a personal computer, but could be used for tiny networks such as ringdemo.
On a supercomputer, after creating a submission jobscript and moving to the model repository directory, enter:
$ sbatch ./jobscripts/MyTestRun.sh # OR $ qsub ./jobscripts/MyTestRun.sh
According to the batch queueing software used by the computer. Or, use SimTracker to create the jobscript, export the script, code, and parameter sets to the supercomputer, and submit the job request to the queue.
You may wish to set...
export N=samplesample export NEURONHOME=samplesample export PYTHONPATH=samplesample
The result files output from each run include certain files that are always written as well as optional results files:
To run this code with different parameters:
#!/bin/bash nrngui [-c "Parameter=Value"] [-c "strdef StringParam" "StringParam=\"StringValue\""] ...
To run a network clamp simulation:
#!/bin/bash nrngui -nopython -c gidOI=21310 -c cellindOI=6 -c stimflag=0 -c "strdef runname" -c runname="\"ca1_1x_exc_08\"" -c "strdef origRunName" -c origRunName="\"ca1_1x_exc_08\"" -c "strdef celltype2use" -c celltype2use="\"poolosyncell\"" -c "strdef resultsfolder" -c resultsfolder="\"00001\"" ./networkclamp_results/ca1_1x_exc_08/00001/run.hoc -c "quit()"
To customize this code...
The blah blah blah.
Please use the following support channels:
- GitHub issues for bug reports and feature requests.
- Our NEURON forum thread to ask questions, discuss features, and for general discussion.
ca1 uses NEURON with HOC:
ca1 can be used with:
- SimTracker — Simulation management tool
- Neuroscience Gateway — user-friendly parallel computing for large scale neural networks
- ModelDB — ModelDB contains many NEURON models and other neuroscience simulator models for various networks and cell types.
Please cite this code if you use it or derive a model from it, using the most appropriate citation below:
- Our ModelDB entry: Cite Bezaire et al (2016), ModelDB (accession number 187604), and McDougal et al (2017). See these instructions for further details.
- Our Open Source Brain entry: Cite Bezaire et al (2016) and see this link for more information.
CC0 1.0 Universal (CC0 1.0): Public Domain Dedication
For more information, see the CC0 Description.
Marianne Bezaire created the ca1 model with help from:
- Ivan Raikov
- Padraig Gleeson
- Andras Ecker
- Kelly Burk
- Michael Hines
- Ted Carnevale
- Ivan Soltesz