action-potential-models

Action potential model example

A typical method of fitting cardiac action potential models is by scaling/updating the maximum conductance parameters of an existing action potential model. This method usually assumes that the underlying kinetics of each ion channel current within the candidate model are correct and perfect (i.e. they match perfectly to the ground truth). Here, we explicitly impose that the underlying kinetics of each ion channel current are imperfect. Then we will ignore the discrepancy and proceed with our analysis with the incorrect assumption (i.e. without acknowledging model discrepancy). Finally, we compare the prediction of those calibrated candidate models under our context of use (COU) to the ground truth (model).

Models

• Model Fink 2008 (candidate model): Fink et al. 2008 model.
• Model TNNP 2004 (ground truth model): ten Tusscher et al. 2004 model.

The differences between the two models can be viewed by running plot-differences.py.

Use of protocol

In this tutorial, we split our protocols into calibration and COU. Note that here we do not have validation protocol, as to emphasise the importance of validation.

• Protocol stim1hz: calibration.
• Protocol randstim: calibration.
• Protocol stim2hz: COU prediction.
• Protocol hergblock: COU prediction.
• Protocol current: COU prediction.

Arguments

• [which_model] can be one of tnnp-2004-w, fink-2008.
• [which_data] and [which_calibration] can be one of stim1hz, randstim.
• [which_predict] can be one of stim1hz, randstim, stim2hz, hergblock, current.

Run the tutorial

1. Run generate-data.py to generate synthetic data with iid Gaussian noise (create data).
2. Run fit.py with arguments [which_model] and [which_data] to calibrate the specified model with the specified (protocol) data from ./data. Alternatively run fit-all.sh.
3. Run predict.py with arguments [which_model], [which_calibration] and [which_predict] to predict the specified (protocol) data (in ./data) with the specified model and calibrated model parameters (in ./out). Alternatively run predict-all.sh.
4. Run mcmc.py with arguments [which_model] and [which_data] to run MCMC for the specified model with the specified (protocol) data from ./data. Alternatively run mcmc-all.sh.
5. Run posterior.py with arguments [which_model], [which_calibration] and [which_predict] to create posterior predictives for the specified (protocol) data (in ./data) with the specified model and MCMC samples of the model parameters (in ./out). Alternatively run posterior-all.sh.

Output

• out: Fitting output etc.
• fig: Output generated figures.
• data: Data generated from the ground truth model.

Utilities

• method: Contains all the useful methods/functions for this tutorial.
• mmt-model-files: action potential models in Myokit mmt format.

Tests

• test-models.py: Simple test for default model setting and forward model simulations.