gletort/physiboss repository overview

⁠PhysiBoSS

PhysiBoSS is a multiscale software to allow for computation of large multi-cellular system, integrating both individual signalling pathway and agent-based representation of cells.

Here we described how to use PhysiBoSS docker image, assuming that you have already read PhysiBoSS GitHub repository⁠ description on what is PhysiBoSS and how to use it.

⁠How to use PhysiBoSS docker image

If PhysiBoSS cannot be compiled on your operating system, it is possible to use PhysiBoSS docker image⁠. For this, you need Docker engine to be installed on your computer. Docker allows you to run an image of a software/environment build in a given operating system with all necessary dependencies, from any operating system. We built the image with a CentOS distribution in which we installed and compiled PhysiBoSS. Thus when you have copied PhysiBoSS docker image, it is ready to be used in a simulation directly. To launch PhysiBoSS Docker image from a Terminal, the command is:

docker run -d -it gletort/physiboss

However, be aware that if you run docker in that way, simulations results will be written on your docker image and not on your local folders. To run it on your local folders, see below.

⁠Running a simulation with docking.sh script

You can find in PhysiBoSS GitHub repository in the Docker folder a bash script to use the docker image directly. You need to give as parameter to this script the full path to the main folder on your computer of the simulation that you want to run, and as second parameter the name of the subfolder containing the parameter file. It will output the results in the folders output and microutput of this folder, as explained in PhysiBoSS description.

For example, to run with docker the Hello World Example from this repository, that you have copied on your machine:

./docking.sh /path/to/Hello/World/Example/folder run0

/path/to/Hello/World/Example/folder is the full path to your simulation folder on your local machine, which must contain the simulation input files.

⁠Running a simulation without the script

To install/update the docker image of PhysiBoSS, you have to pull it on your machine:

docker pull gletort/physiboss

To run a simulation, you could run it directly inside the Docker image, but you don't access the files directly then. To link it to your local folders, you have to mount the local directory in which you have prepare the simulation. Outputs will then also be written in that directory. The command is the following:

docker run -d -it -v /path/to/local/directory:/home/PhysiBoSS/running/name_of_simulation gletort/physiboss sh -c "cd /home/PhysiBoSS/running/name_of_simulation/subfolder_of_simulation_to_run; /home/PhysiBoSS/bin/PhysiBoSS > docking.txt"

With:

• -d to run the docker image in background
• -v path1:path2 allows you to mount the local directory path1 and placed it in PhysiBoSS image as directory path2. path1 is thus the path to the folder in your computer where the simulation to run is, and where the results will be written during the simulation (in the output folder). path2 is a "virtual" path to the simulation inside the docker image. You have to precise the full path to the directories. NB: In the PhysiBoSS image, we created a folder /home/PhysiBoSS/running in which to place all running simulations. But of course you can specified another path.
• gletort/physiboss precises the image that you want to run
• sh -c "cmd" allows you to define which commands you want to run in the image. Here, you move to the simulation folder (cd /path/), in the subfolder that you want to run. From there, you call PhysiBoSS executable and run the simulation (/home/PhysiBoSS/bin/PhysiBoSS). And we redirect all the printed messages indicating the simulation evolution in the file docking.txt that you can check to see to which point the simulations is.