Docker container for OpenCOR

This Docker container can be used to run a simulation of a CellML / SED-ML file using OpenCOR through Python.

You can either build the container yourself or use one of the prebuilt versions available on Docker Hub at https://hub.docker.com/repository/docker/opencor/opencor.

A Docker image to execute CellML/SED-ML based COMBINE archives using OpenCOR is available from BioSimulators.

Pull a prebuilt version of the container

docker pull opencor/opencor[:<tag>]

where <tag> is the tag of a specific version (yyyy-mm-dd) or the latest version/snapshot (latest). If no tag is provided then the latest version/snapshot is assumed. (Only stable snapshots are used.)

Build the container yourself

From GitHub

Default version

A default container can be built using:

docker build -t opencor https://github.com/opencor/docker.git

Specific version

A container with a specific version of OpenCOR can be built using:

docker build -t opencor --build-arg version=<version> https://github.com/opencor/docker.git

where <version> can be either x.y[.z] (for an official version of OpenCOR) or yyyy-mm-dd (for a snapshot). Note that there is currently no official version of OpenCOR with Python support.

From a clone

If you clone the GitHub repository, you will not only be able to build a container with a default or a specific version of OpenCOR, but also a container based on a local archive of OpenCOR:

docker build -t opencor --build-arg archive=<archive> <dir>

where <archive> is the filename of a .tar.gz file located in <dir>, the directory where the GitHub directory was cloned.

Run the container

Default

docker run <image> <url>

where <image> is the name of the image (i.e. opencor/opencor if you pulled a prebuilt version of the container or opencor if you built it yourself), and <url> is the URL of a CellML / SED-ML file, e.g. https://models.physiomeproject.org/workspace/5f5/rawfile/125f548ce204c1d815298d2c8c1d9b774d89e3a7/mcintyre_richardson_grill_model_2001.sedml.

This will run the CellML / SED-ML file and output the results for the variable of integration, as well as all the state variables in the model, this in a JSON format. For example, for the above URL, we get:

{
  "environment/t": [
    0.0,
    ...,
    100.0
  ],
  "membrane/V_m": [
    -88.5901439103062,
    ...,
    -89.42705739292656
  ],
  "fast_sodium_channel/fast_sodium_channel_m_gate/m": [
    0.0302964457761589,
    ...,
    0.027969909786511916
  ],
  "fast_sodium_channel/fast_sodium_channel_h_gate/h": [
    0.841520865130776,
    ...,
    0.8560318033219113
  ],
  "persistent_sodium_channel/persistent_sodium_channel_p_gate/p": [
    0.0969864645712442,
    ...,
    0.11512881732943588
  ],
  "slow_potassium_channel/slow_potassium_channel_s_gate/s": [
    0.00997371545602793,
    ...,
    0.1297692977394969
  ],
  "juxtaparanodal_fast_potassium_channel/juxtaparanodal_fast_potassium_channel_n_gate/n": [
    0.000886041197111556,
    ...,
    0.000463917000795369
  ]
}

Using a configuration file

docker run -i <image> <url> < <config>

where <config> is the filename of a configuration file in the JSON format, e.g. config.json:

{
  "simulation": {
    "Starting point": 5,
    "Ending point": 50,
    "Point interval": 0.5
  },
  "parameters": {
    "fast_sodium_channel/g_Naf": 3.5,
    "membrane/V_m": -35
  },
  "output": [
    "fast_sodium_channel/i_Naf",
    "leakage_channel/i_Lk",
    "membrane/V_m",
    "membrane/V_m/prime",
    "persistent_sodium_channel/i_Nap",
    "slow_potassium_channel/i_Ks"
  ]
}

The configuration file is used to configure the simulation (using the information in the simulation section, if present), customise the initial state of the model (using the information in the parameters section, if present). The simulation is then run and the variable of integration, as well as all the variables listed in the output section (if present otherwise all the state variables), are outputted. For example, for https://models.physiomeproject.org/workspace/5f5/rawfile/125f548ce204c1d815298d2c8c1d9b774d89e3a7/mcintyre_richardson_grill_model_2001.sedml and the above configuration file, we get:

{
  "environment/t": [
    5.0,
    ...,
    50.0
  ],
  "membrane/V_m": [
    -35.0,
    ...,
    -72.4735314003255
  ],
  "membrane/V_m/prime": [
    -210.57350734894032,
    ...,
    3.572051321498071
  ],
  "fast_sodium_channel/i_Naf": [
    -0.006961885968193237,
    ...,
    -0.3051159365312024
  ],
  "leakage_channel/i_Lk": [
    0.385,
    ...,
    0.12268528019772151
  ],
  "persistent_sodium_channel/i_Nap": [
    -0.0007754473404489828,
    ...,
    -0.701626253584342
  ],
  "slow_potassium_channel/i_Ks": [
    0.04388434800652289,
    ...,
    0.8769128072748268
  ]
}

Note that rate variables have an identifier that ends with /prime, e.g. membrane/V_m/prime refers to the rate variable for the membrane/V_m state variable.

Use the BioSimulators Docker image

The BioSimulators Docker image can be obtained using:

docker pull ghcr.io/biosimulators/opencor

The image can be used to execute CellML/SED-ML based COMBINE archives using:

docker run -it --rm \
       --mount type=bind,source=/path/to/in/dir,target=/tmp/in \
       --mount type=bind,source=/path/to/out/dir,target=/tmp/out \
       ghcr.io/biosimulators/opencor \
       -i /tmp/in/archive.omex \
       -o /tmp/out

All reports and plots described in the SED-ML files in the archive will be saved in HDF5 and PDF formats. More information about these formats is available at https://biosimulators.org/conventions/simulation-reports.