Skip to main content

Data from: In vitro to in vivo extrapolation from three-dimensional hiPSC-derived cardiac microtissues and physiologically based pharmacokinetic modeling to inform next-generation arrythmia risk assessment

Cite this dataset

Daley, Mark et al. (2024). Data from: In vitro to in vivo extrapolation from three-dimensional hiPSC-derived cardiac microtissues and physiologically based pharmacokinetic modeling to inform next-generation arrythmia risk assessment [Dataset]. Dryad.


Proarrhythmic cardiotoxicity remains a substantial barrier to drug development as well as a major global health challenge. In vitro human pluripotent stem cell-based new approach methodologies have been increasingly proposed and employed as alternatives to existing in vitro and in vivo models that do not accurately recapitulate human cardiac electrophysiology or cardiotoxicity risk. In this study, we expanded the capacity of our previously established three-dimensional human cardiac microtissue model to perform quantitative risk assessment by combining it with a physiologically based pharmacokinetic model, allowing a direct comparison of potentially harmful concentrations predicted in vitro to in vivo therapeutic levels. This approach enabled the measurement of concentration responses and margins of exposure for two physiologically relevant metrics of proarrhythmic risk (i.e., action potential duration and triangulation assessed by optical mapping) across concentrations spanning three orders of magnitude. The combination of both metrics enabled accurate proarrhythmic risk assessment of four compounds with a range of known proarrhythmic risk profiles (i.e., quinidine, cisapride, ranolazine, and verapamil) and demonstrated close agreement with their known clinical effects. Action potential triangulation was found to be a more sensitive metric for predicting proarrhythmic risk associated with the primary mechanism of concern for pharmaceutical-induced fatal ventricular arrhythmias, delayed cardiac repolarization due to inhibition of the rapid delayed rectifier potassium channel, or hERG channel. This study advances human induced pluripotent stem cell-based three-dimensional cardiac tissue models as new approach methodologies that enable in vitro proarrhythmic risk assessment with high precision of quantitative metrics for understanding clinically relevant cardiotoxicity.

README: Data from: In vitro to in vivo extrapolation from three-dimensional hiPSC-derived cardiac microtissues and physiologically based pharmacokinetic modeling to inform next-generation arrythmia risk assessment



Description of the data and file structure

Data Collection:

Engineered cardiac microtissues generated from human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) were loaded with the voltage sensitive dye, Di4-ANEPPS, and imaged at 37 degrees Celsius during 0.5 Hz field stimulation to record action potentials across a two-second interval. Action potentials were recorded for each microtissue across six half-log-spaced doses of one of four drugs (quinidine, cisapride, ranolazine, or verapamil) after 15-minute exposures as well as at baseline. For each action potential, action potential duration at maximum rate of repolarization (APDmxr), a measure of total action potential length, and action potential triangulation (APDtri), a measure of repolarization, were quantified. Microtissues were generated in batches and were imaged together in molds of 35 microtissues each. Response data is averaged across all recorded beats for each mold and microtissue. Additional details are available in the full manuscript (

Data Analysis:

Included analysis calculates the point of departure (POD) from concentration-response curves of hiPSC-derived cardiac microtissues in response to cardiotoxic drugs (quinidine, cisapride, ranolazine, and verapamil). The POD is the lowest concentration determined to cause a deviation from baseline behavior and is used to define the lowest potentially toxic dose from concentration-response data. In this algorithm, POD is calculated by resampling equally weighted data using bootstrapping and fitting the resampled data with 1) cubic spline and 2) sigmoidal function and then finding the point at which the dose response curve crosses a threshold 5% above baseline. The script analyzing this data set,, is available from Zenodo ( Run this script in the same directory as the included .txt files and follow the instructions below to calculate and visualize POD confidence intervals.

Data Filename Description:

Filenames are given in the format of METRIC_NORMALIZATION_DRUGNAME.txt.


  • APDmxr = Action potential duration (APD) at maximum rate of repolarization
  • APDtri = Action potential triangulation (APDmxr – APD at 50% repolarization)


  • Raw = Absolute action potential metric data (in milliseconds)
  • Norm = Action potential metric data is normalized by each microtissue's baseline (given as a unitless fraction).


  • Quinidine
  • Cisapride
  • Ranolazine
  • Verapamil

For example, the file APDmxr_Norm_Cisapride.txt was generated from microtissues treated with cisapride and includes APDmxr data normalized by each microtissue's baseline.

Description of data file column names:

Mold: Identifying number of mold from which data was recorded (not de-duplicated across drugs)

Microtissue: Identifying number of microtissue from which data was recorded (not de-duplicated across drugs)

Concentration: Concentration of compound during data collection (micromolar)

Value: Mean action potential metric data (APDmxr or APDtri) (milliseconds)

Stdev: Beat to beat standard deviation of Value

NumBeats: Number of action potentials analyzed to obtain Value (Up to 4 total).

Data Files:



















Python version 3.11.7

Anaconda Conda version 29.9.0

From the Anaconda Powershell prompt, please install following python packages:

  • Numpy
  • SciPy
  • Pandas
  • MatPlotLib
  • Seaborn

To Use:

1)    Download the “SupplementaryMaterial_PoDCode_SampleData” zipped folder

2)    Open the Python file “” from the folder with the sample data

3)    Run the Python file.

4)    Type “PoD_CI_Est_equalweight()” into command line and hit “return”

5)    Program will output each data filename loaded (APDmxr, APDtri) and output mean PoD (confidence intervals) and mean adjusted PoD (confidence intervals) below the filenames. Also, kernel density estimation plots for each compound are displayed.

License and Copyright

Copyright © 2024, Bum-Rak Choi, Mark Daley, and Peter Bronk. This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. For a copy of the GNU General Public License, see


National Institute of Environmental Health Sciences, Award: U01 ES028184

Brown University, Brown Biomedical Innovations to Impact