PyPNS is a Python module for the simulation of peripheral nerves. Axon models are simulated in the compartmental simulator NEURON over its Python interface. Extracellular potentials from membrane currents or stimulation electrodes are calculated in a resistive, electro-quasistatic approximation of the Maxwell equations from either a homogeneous analytical solution, precomputed and imported finite element model outputs or analytical functions fit to FEM results.
PyPNS was developed in the Department of Bioengineering, Centre of Neurotechnology at Imperial College London.
This scientific software is released under the GNU Public License GPLv3.
For a quick start the Binder docker can be used that was contributed by russelljjarvis at fun-zoological-computing.
Click on the picture below to start the docker in mybinder.org. Type code like in the picture or launch work/PyPNS/jupyter_test.ipynb.
To install PyPNS you will need:
- Python3
- Python modules numpy, scipy and matplotlib
- NEURON (from http://www.neuron.yale.edu) compiled as a Python module. IMPORTANT: Do not rely on
pipfor the installation of NEURON. Below is a guide on how to install NEURON including the Python interface from source. PyPNS is tested for NEURON 7.6.7 with InterViews 19
It is advisable to install NEURON from source. The following description is closely based on Andrew Davidson's tutorial and copied here to make sure it is available. Another good description can be found on the official page here.
Download the NEURON source nrn-nn.tar.gz
and the corresponding version of InterViews iv-mm.tar.gz from here.
Replace mm with your InterViews version number.
N=`pwd`
tar xzf iv-mm.tar.gz
cd iv-mm
./configure --prefix=`pwd`
make
make install
Replace nn with your NEURON version number and mm with your InterViews version number.
cd ..
tar xzf nrn-nn.tar.gz
cd nrn-nn
./configure --prefix=`pwd` --with-iv=$N/iv-mm --with-nrnpython
make
make install
If you want to run parallel NEURON, add --with-paranrn to the configure options. On Mac OS X, I have found I need to add PYLIB=-lpython PYLIBLINK=-lpython to the configure line.
Replace nn with your NEURON version number and x86_64 with your architecture code.
export PATH=$N/nrn-nn/x86_64/bin:$PATH
cd src/nrnpython
python setup.py install
This command (which will probably have to be run as root or using sudo) will install the neuron package to your site-packages directory. An alternative, especially if you don't have root privileges, is:
python setup.py install --prefix=~
which will install the neuron package to ~/lib/python/site-packages. You will then have to add this directory to the PYTHONPATH environment variable:
export PYTHONPATH=$PYTHONPATH:~/lib/python/site-packages
Using pip or easy install
pip install PyPNS
easy_install PyPNS
From source:
tar -xzf PyPNS-x.x.tar.gz
cd PyPNS-x.x
python setup.py install
To run PyPNS properly, several NEURON extensions need to be compiled for the myelinated axon model. Those are located in the mods-directory of PyPNS. Download this directory and run
nrnivmodl
from the console. A new folder will be generated containing the compiled files. They need to be present in the working directory of your project.
If FEM results are to be used for either recording or stimulation, the 'Fields'-directory needs to be present within the working directory as well. Different field dictionaries need to be contained within subdirectories. Subdirectory name equals field name. One example field can be downloaded from GitHub.
For a quick first test, download the test.py script from GitHub and run it:
python test.py
To reproduce figures of the paper go to this figure-repository.
For further description of the different components of PyPNS, compile the Sphinx documentation. To do so, navigate to documentation/source within the PyPNS directory and run
sphinx-build -b html . ../HTML