Atrial Low-K code

This repository contains the code needed to reproduce the figures and simulations for the paper "Immediate and delayed response of simulated human atrial myocytes to clinically-relevant hypokalemia" 10.3389/fphys.2021.651162. To cite this work, please use the information provided in the CITATION file.

Requirements

• Python 3.4+
• Myokit 1.30.1+
• A working OpenCL installation is required for the tissue simulations
• Installation help for Myokit and OpenCL

Shared code and models

• Shared code is stored in shared.py and used by several figures.

Models:

• The Courtemanche et al. 1998 model. Original paper doi: 10.1152/ajpheart.1998.275.1.h301. Myokit model code was adapted from an existing CellML model in PMR.
• The Grandi, Pandit, Voigt et al. 2011 model. Original paper doi: 10.1161/CIRCRESAHA.111.253955 Myokit model code was newly implemented and a CellML model was uploaded to PMR.
• The Koivumäki, Korhonen et al. 2011 model. Original paper doi: 10.1371/journal.pcbi.1001067 Myokit model code was newly implemented and a CellML model was uploaded to PMR.
• The Maleckar et al. 2008 model. Original paper doi: 10.1016/j.pbiomolbio.2009.01.010 Myokit model code was adapted from an existing CellML model in PMR.
• The Ni et al. 2017 model. Original paper doi: 10.3389/fphys.2017.00946 Myokit model code was newly implemented and a CellML model was uploaded to PMR.
• The Nygren et al. 1998 model. Original paper doi: 10.1161/01.RES.82.1.63 Myokit model code was adapted from an existing CellML model.
• The Voigt, Heijman et al. 2013 model. Original paper doi: 10.1016/j.yjmcc.2013.03.011 Myokit model code was adapted from this code and a CellML model was uploaded to PMR.

Main figures

To re-run the single cell simulations, and create the main figures, use:

• Figure 1: f1.py
• Figure 2: f2.py
• Figure 3: f3.py
• Figure 4: f4.py
• Figure 5: f5.py
• Finally, to generate the non-schematic parts of Figure 6, use f6.py.

These store their figures in directories figure-1, figure-2, etc.

Supplementary figures

Several supplementary figures are provided:

• s1-vr.py creates supplementary figure 1, showing digitised data from Gelband et al. 1972 and Ten Eick & Singer, 1979, and comparing with model predictions.
• s2-ap-cat.py creates a model comparison figure showing AP and CaT.
• s3-relative-contributions.py creates a figure showing relative contributions of all currents.
• s4-restitution.py <model> creates a restitution curve for the given model, and plots it alongside data stored in literature-data, digitised from Van Wagoner et al..
• s5-ik1-ko.py generates a figure showing the three ways in which external K+ affects IK1.
• s6-ik1-x.py generates a figure showing the K-dependence of the rectifiation variable in four models.
• s7-fast-slow.py <model> generates a figure showing short and long-term changes in concentrations, AP, IK1, INaK, and INaCa.
• s8-strand.py performs strand simulations and creates an overview figure.
• s9-dvdtmax.py <model> generates a figure showing maximum dV/dt for all models.
• s10-ina-sd.py <model> generates a figure with single-cell excitability data (INa availability and SD curves).
• s11a-rm-calculations.py creates a figure showing details of membrane resistance calculations, as do
• s11b-rm-ko.py, and
• s11c-rm-low.py (which can be customised by changing the variables k, dt, and `dv`` inside the script).
• s11d-rm.py creates a membrane resistance figure for a chosen model.
• s12-iv.py <model> generates a figure showing instantaneous IV relationships at low K.

Linux or OS/X users can use the all-x scripts to generate figures for all models. Similarly, the all script regenerates all figures in this study.