A Python package implementing computational models of the globular bushy cells (GBCs) in the mammalian ventral cochlear nucleus. Phenomenological models of endbulbs of Held with with short term synaptic depressions are also included.
Look for examples in the examples directory (e.g. nrn_gbc_demo.py and nrn_gbc_demo_with_default_weights.py). Note that some examples require cochlea to be installed.
Install the dependencies:
- Python 2.7
- Numpy
- NEURON (with Python bindings): http://neuron.yale.edu/neuron/
- Brian: http://briansimulator.org/
Clone the git repo with:
git clone https://github.com/mrkrd/cochlear_nucleus.git cd cochlear_nucleus
Compile NEURON's models:
make
or manually:
cd cochlear_nucleus/nrn && nrnivmodl cd -
Install the package itself:
python setup.py develop --user
Rudnicki M. and Hemmert W. (2017). High entrainment constrains synaptic depression levels of an in vivo globular bushy cell model. Frontiers in Computational Neuroscience, Frontiers Media SA, 2017, 11, pp. 1-11. doi:10.3389/fncom.2017.00016 https://www.frontiersin.org/articles/10.3389/fncom.2017.00016/full
BibTeX entry:
@Article{Rudnicki2017, author = {Marek Rudnicki and Werner Hemmert}, title = {High Entrainment Constrains Synaptic Depression Levels of an In vivo Globular Bushy Cell Model}, journal = {Frontiers in Computational Neuroscience}, year = {2017}, volume = {11}, pages = {1--11}, month = {mar}, doi = {10.3389/fncom.2017.00016}, publisher = {Frontiers Media {SA}}, url = {https://www.frontiersin.org/articles/10.3389/fncom.2017.00016/full}, }
This work was supported by the German Research Foundation (DFG) within the Priority Program “Ultrafast and temporally precise information processing: normal and dysfunctional hearing” SPP 1608 (HE6713), the Technische Universität München within the funding programme Open Access Publishing, and the German Federal Ministry of Education and Research within the Munich Bernstein Center for Computational Neuroscience (reference number 01GQ1004B).
The project is licensed under the GNU General Public License v3 or later (GPLv3+).