Calibrate cardiac electrophysiology cell models using Approximate Bayesian Computation with the myokit and pyabc libraries.
- Create local clone of this repository.
- Recommended Create new python environment using miniconda using the provided
environment.ymlfile. e.g.conda env update -f environment.yml - Activate the previously created miniconda environment. e.g.
conda activate ionchannelABC - Install CVODE dependency required by myokit as described here.
- Update the paths to a SUNDIALS installation in the
myokit.inifile which is created on the user's home path~/.config/myokit/myokit.ini. - If not already existing, create an environment variable to a temporary directory
TMPDIRnecessary for myokit to save local files. A line could be added to your~/.bashrc:export TMPDIR=/path/to/tmp/directory. - Finally, install the ion-channel-ABC package by navigating to your cloned repository and running
python setup.py install.
Example Jupyter notebooks demonstrating use of key features are available in (docs/examples) folder. It is recommended to start with the getting_started.ipynb notebook.
Update: It is suggested to start by looking at notebooks in the docs/examples/human-atrial folder which are more current (e.g. nygren_ina_original.ipynb) while the getting_start.ipynb notebook is updated.