ion-channel-ABC

Calibrate cardiac electrophysiology cell models using Approximate Bayesian Computation with the myokit and pyabc libraries.

Installation

1. Create local clone of this repository.
2. Recommended Create new python environment using miniconda using the provided environment.yml file. e.g. conda env update -f environment.yml
3. Activate the previously created miniconda environment. e.g. conda activate ionchannelABC
4. Install CVODE dependency required by myokit as described here.
5. Update the paths to a SUNDIALS installation in the myokit.ini file which is created on the user's home path ~/.config/myokit/myokit.ini.
6. If not already existing, create an environment variable to a temporary directory TMPDIR necessary for myokit to save local files. A line could be added to your ~/.bashrc: export TMPDIR=/path/to/tmp/directory.
7. Finally, install the ion-channel-ABC package by navigating to your cloned repository and running python setup.py install.

Running

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.