Cardiac TdP Risk Stratification Modelling of Anti-Infective Compounds including Chloroquine and Hydroxychloroquine

Code associated with a Royal Society Open Science paper by Dominic G. Whittaker, Rebecca A. Capel, Maurice Hendrix, Xin Hui S. Chan, Neil Herring, Nicholas J. White, Gary R. Mirams, Rebecca-Ann B. Burton.

Requirements

Processing and plotting the simulation data in this repository requires installation of python 3 and certain libraries within. We recommend installing packages and running the scripts in a virtual environment to avoid version conflicts. In order to do this, follow these steps:

• virtualenv folder_name (or virtualenv --python=python3 folder_name if you have both python 2 and 3). If virtualenv is not recognised you may need to call it as python -m virtualenv folder_name or (python -m virtualenv folder_name). If that doesn't work you may need to install virtualenv first with pip install virtualenv.
• activate the virtual environment with source folder_name/bin/activate (or folder_name/Scripts\activate on Windows)
• now get the source code from git: git clone https://github.com/CardiacModelling/risk-stratification-anti-malarials.git
• install the required packages by typing pip install -r requirements.txt

Generating the data

The simulations in this paper use the ApPredict tool. In order to facility reproducibility we have published the exact version of the tool used to Docker Hub. In order to generate the data used in the paper, install Docker. If you are on Windows you may have to set Docker to use Linux containers. After installing Docker run the following command:

• docker run -it cardiacmodelling/appredict-in-papers:brute_force_0.0.1 /bin/bash

The first time you run this, this will download the relevant Docker image and it will log into a virtual environment where you can run the following commands:

• cd apps/ApPredict/

We also recommend activating a screen session if using Linux as it may take many hours to generate the simulation data, which are produced by typing a command which tells ApPredict which ion channels to block and by which amount for each drug/combination. For example, the data for hydroxychloroquine are generated by typing

• ./ApPredict.sh --model 8 --pacing-freq 0.5 --pic50-herg 5.25 --pic50-spread-herg 0.139 --pic50-cal 4.57 --pic50-spread-cal 0.181 --pic50-iks 5.03 --pic50-spread-iks 0.127 --plasma-conc-high 100 --plasma-conc-count 19 --plasma-conc-logscale True --no-downsampling True --credible-intervals 60 70 80 90 95 --brute-force 1000 --output-dir HCQ &> testoutput/HCQ.txt &

which runs the simulation in detached mode and stores the data in testoutput/HCQ/ and console output in testoutput/HCQ.txt. For a full list of the commands used to generate simulation data for all drugs and combinations see commands. Once the data have been generated, inside the directory in which you wish to store the data (outside the Docker container) run:

• docker container list and note the CONTAINER ID
• docker cp CONTAINER ID:/home/appredict/apps/ApPredict/testoutput .

For convenience we have already stored all of the simulation data in testoutput.

Plotting the figures

In order to generate Figure 1 from the paper, simply type:

• python Figure1.py (or python3 Figure1.py if you have both python 2 and 3 installed).
• Figures 2 and supplementary Figures S1 and S2 can be rendered using the same command (i.e. python Figure2.py, python FigureS1.py and python FigureS2.py).
• Pre-generated and saved figures used in the paper can be found in Figures/.

Acknowledging this work

If you publish any work based on the contents of this repository please cite (CITATION file):

Whittaker, D. G., Capel, R. A., Hendrix, M., Chan, X. H. S., Herring, N., White, N. J., Mirams, G. R., Burton, R. A. B. (2021). Cardiac TdP Risk Stratification Modelling of Anti-Infective Compounds including Chloroquine and Hydroxychloroquine. Royal Society Open Science, 8:210235.