Logistic regression analysis of populations of electrophysiological models to assess proarrythmic risk.

This package contains the code used to investigate proarrythmic mechanisms on a population of 
electrophysiological models with the method of multivariable logistic regression.
In this example, the Morotti et al. model of human atrial myocyte (J Mol Cell Cardiol. 2016 Jul;96:63-71,
available for download on this website) is used to create a family of 1000 model variants by perturbing
model parameters, and to simulate an electrophysiological protocol that enhances EAD proclivity. 
Logistic regression analysis allows relating changes in model parameters to the presence/absence of EADs.



                                      Matlab code

Contents:
__________________________________________________________________________________________________________

                                      MAIN FILES
__________________________________________________________________________________________________________
1) SA_generate_parameters.m
File for generation of random perturbations of model parameters.
__________________________________________________________________________________________________________
2) SA_analyze_EAD.m
File for cyclic execution of EAD protocol and analysis of model outputs. It uses the matrix created with
SA_generate_parameters.m.
__________________________________________________________________________________________________________
3) SA_logistic.m
File for logistic regression analysis and plot of the results. This file uses the matrices created with
SA_generate_parameters.m and SA_analyze_EAD.m.
__________________________________________________________________________________________________________

                                      SUPPORTING FILES
__________________________________________________________________________________________________________
4) morotti_et_al_ham_ina_ran_model_SA.m
ODE model of excitation-contraction coupling in the human atrial cell used in SA_analyze_EAD.m. This file
was modified from the one published in Morotti et al. J Mol Cell Cardiol. 2016 Jul;96:63-71. for allowing
modulation of model parameters).
__________________________________________________________________________________________________________
5) EAD_occurrence.m
Function called by SA_analyze_EAD.m for the assessment of EAD occurrence.
__________________________________________________________________________________________________________
6) rotateXLabels.m
Function called by SA_logistic.m for rotating the labels on the X axis in some figures.
__________________________________________________________________________________________________________

                                      MAT FILES
__________________________________________________________________________________________________________
7) yf_ham_ina_ran_ACh0p1_1Hz.mat
Initial conditions (obtained with the baseline model at 1-Hz pacing in presence of 0.1 uM acetylcholine)
used for simulations in SA_analyze_EAD.m.
__________________________________________________________________________________________________________
8) SA_par_matrix_1000_s0p1.mat
Matrix of the scaling factors used in the example shown in this paper (this file was created and saved
with the SA_generate_parameters.m, and then used in SA_analyze_EAD.m and SA_logistic.m).
__________________________________________________________________________________________________________
9) SA_outputs_matrix_1000_s0p1.mat
Matrix of the model outputs obtained in the example shown in the paper (this file was created and saved 
with SA_analyze_EAD.m, and then used in SA_logistic.m).
__________________________________________________________________________________________________________


Reference:
Morotti S & Grandi E.
Logistic regression analysis of populations of electrophysiological models to assess proarrythmic risk.
MethodsX. 2016 Dec 23; 4:25-34. doi: http://dx.doi.org/10.1016/j.mex.2016.12.002

Please cite the above paper when using this code.