README

INFORMATION ABOUT THE RELATED PUBLICATION

Gibbs et al. (2026) Accelerated intrinsic beating rate in heterogeneously coupled human pluripotent stem cell-derived cardiomyocytes can underlie focal ventricular tachycardia in regenerative therapy

Link to published article: https://doi.org/10.1016/j.premed.2026.100035

Please contact Dr. Patrick M. Boyle (pmjboyle@uw.edu) with any questions, concerns, complaints, or difficulties.

This dataset contains data required to reproduce computational simulations presented in the 2026 paper by Gibbs et al. (JPMHD 10.1016/j.premed.2026.100035). Please consider citing the original paper as well as this repository if you reuse, adapt, or extend any of the models or simulation protocols described below.

Abstract from paper:

Cardiac stem cell therapies seek to replace damaged tissue following myocardial infarction to prevent heart failure. Human pluripotent stem-cell derived cardiomyocyte (hPSC-CM) injection can be used to remuscularize the heart and improve function but also cause ventricular tachycardia (VT). Electrical mapping studies suggest this VT is focal (i.e., driven by spontaneous excitations), but true mechanisms remain unknown. Computational approaches may provide new insights into these arrhythmias, but current simulation frameworks have not replicated focal VT. In this study, we conduct simulations with biophysically plausible representations of cell injection and the formation of gap junctions between hPSC-CM grafts and host tissue. We examine hPSC-CM ionic models with baseline (1.1 Hz) and accelerated (2.5, 4 Hz) beating rates and assess whether these conditions can produce focal VT that is adequately rapid and robust to overpower sinus rhythm. For simulations conducted in human ventricular slice geometries with non-human primate graft patterns, we assessed susceptibility to focal VT. For the 1.1 Hz model, focal VT was only seen at bradycardic heart rates (40bpm); at faster heart rates (60 or 80bpm), ectopic activity from the graft could not out-compete Purkinje-like endocardial excitations. In contrast, focal VT was seen for all tested heart rates when faster ionic models were used. Doubling graft size to cover larger infarct areas increased focal VT susceptibility in all cases. We thus conclude that post-injection increase in hPSC-CM spontaneous beating rate is a potential mechanism for focal VT; experimental validation of this hypothesis could lead to safer cell injection schemes.

Description of the data and file structure

Supplemental Videos referenced in the paper can be found under the following names:

• Supplemental_Video_1.mp4
• Supplemental_Video_2.mp4
• Supplemental_Video_3.mp4

Raw data for Figure 6 can be found in Figure6_1Xvs2X.csv.

The Geometries.tar archive contains the finite element meshes for all three human geometries tested and the different graft placements at both 1X scale and 2X. Geometries with grafts applied at 2X are noted as Geometry*_2X . For the 1X graft geometries the naming scheme is as followed GeometryX_GraftYVZ where X=1,2,3, Y=1,2,3, and Z=1,2 and 3 in the case of Geometry3.

The folder structure looks like the follow:
Geometries/
- Geometry1/
- Geometry2/
- Geometry3/
- HumanGraft.mshz

Specific finite element mesh file types found in the Geometries Folder:

• .pts: description of points in Cartesian coordinates (spacial unit: microns)
• .elem: description of how nodes (zero indexed) are interconnected by elements
• .lon: fiber orientations for each element
• .mshz: state files for model visualization

Example on how to visualize models:

meshalyzer Geometry1/Geometry1_Graft1V1.pts HumanGraft.mshz

HeatmapFiles.tar is a compressed archive containing the raw values and jupyter notebook script need to recreate Figure 3 D-F, Figure 4 D-F, and Figure 5 D-F. Each text file contains the raw values fraction of simulations resulting in VT for each graft vs sinus pacing. The naming scheme notes which geometry and model were used.

HeatmapFiles/
- Geometry1_Slow.txt
- Geometry1_Medium.txt
- Geometry1_Fast.txt
- Geometry2_Slow.txt
- Geometry2_Medium.txt
- Geometry2_Fast.txt
- Geometry3_Slow.txt
- Geometry3_Medium.txt
- Geometry3_Fast.txt
- HeatmapPlots.ipynb : Jupyter notebook file to recreate all heatmaps

FindBPM.tar is a compressed archive containing an example of how BPM was calculated based on the local transmural extracellular potential gradients (∇Φet) across the electrode in simulation. Example raw values and the jupyter notebook script used are provided. The .dat files contain time vs potential.

FindBPM/
- FastModel_Geometry2Graft1_Left.dat
- FastModel_Geometry2Graft1_Right.dat
- FindBPM.ipynb
- MediumModel_Geometry3Graft1_Left.dat
- MediumModel_Geometry3Graft1_Right.dat
- SlowModel_Geometry1Graft1_Left.dat
- SlowModel_Geometry1Graft1_Right.dat

Simulations.tar is a compressed archive containing an example model with all files needed to rerun and visualize simulations in openCARP and Meshalyzer (.pts, .lon, .elem).

Simulations/
- FastRate.model
- FastRate.sv
- Geometry1_40bpm.par
- Geometry1_Graft1V1_Results/
- Geometry1_Graft1V1_Results/SlowRate/
- Geometry1_Graft1V1_Split0.800_N7*
- MediumRate.model
- MediumRate.sv
- RunExampleSimulation.sh
- SlowRate.model
- SlowRate.sv
- TT2_BZ.sv
- vm.mshz

The .sv files contain initial conditions for all cell-scale ionic models used in simulation studies.
The file .par provides the baseline parameters used to run all simulations in this study.
The file .model comes contain the modifications made the ionic parameters for each

To compile these file into something openCARP can use for simulations be sure to run make_dynamic_model.sh where you have openCARP installed.
Example: /Software/openCARP/bin/make_dynamic_model.sh SlowRate.model

Rerunning the driver script will produce all of the files shown in Geometry1_Graft1V1_Results/SlowRate subfolder
- electrics.log file include information about the ODE/PDE solvers used
- par_stats.dat file include log information about the parabolic solver
- parameters.par file contain all the options defined for each simulation
- petsc_err_log.txt file contain errors encountered by the PETSc solver.
- protocol.trc file contain information about electrical stimuli applied
- Stimulus_0.trc file contain information about electrical pulse applied to the left side of the mesh
- Stimulus_1.trc file contain information about electrical pulse applied to the right side of the mesh
- vm_act-thresh.dat file contain activation nodes and times in temporal sequence, measured via threshold crossing (-40 mV)
- vm.igb file contains voltage overtime data that can visualized in meshalyzer

Example usage – tested using openCARP v7.0 on Ubuntu 18.04.6

The command below (text following $) can be executed on a properly configured system.

$ bash -u RunExampleSimulation.sh:

Data visualization – tested using meshalyzer v5.2 on Ubuntu 18.04.6

The command below can be executed on a properly configure system to visualize the results of a particular simulation; modify the mesh name provided and the path to the relevant vm.igb file to visualize results of different simulations.

$ meshalyzer Geometry1_Graft1V1_Split0.800_N7. Geometry1_Graft1V1_Results/SlowRate/vm.igb vm.mshz

The first line after the invocation of the visualization program is the path to the .pts file associated with the relevant mesh. The second line is the path to the vm.igb file containing membrane voltage over time. The third line is the path to a meshalyzer state file compatible with v5.2 of the software.

Sharing/access information

All files contained in this repository are published under the Creative Commons CC0 license. Future users are thus allowed and encouraged to distribute, remix, adapt, and build upon the material in any medium or format, with no conditions. Although CC0 does not require attribution, we encourage all parties who reuse this dataset in any way to cite the paper with which it was published

Code/software information

openCARP and meshalyzer information can be found at: https://opencarp.org/

FIN