Installation
This page details how to install OpenMMDL on your local device.
Installation via conda-forge
OpenMMDL is implemented in conda-forge and can be installed for linux-based system with the following command line:
conda install -c conda-forge openmmdl
Installation via repository
OpenMMDL can be installed from the repository by creating the conda environment defined in environment.yml.
The first step of the installation of OpenMMDL on your device consists in the cloning of OpenMMDL to your home directory. This can be achieved by using the following command lines:
cd ~
git clone https://github.com/wolberlab/OpenMMDL
OpenMMDL supports Python 3.11-3.12 and uses several packages, which can be installed in a separate conda environment (we recommend miniconda):
cd OpenMMDL
conda env create -f environment.yml -n openmmdl
After installation, activate the conda environment:
conda activate openmmdl
Install the OpenMMDL package with the required entry points via pip:
pip install .
This should lead to the environment having now all the required packages and entry points for OpenMMDL.
Checking the installation
After installing OpenMMDL, you can check whether the entry points and packages are available with:
openmmdl check
This command prints the OpenMMDL version, Python version, platform information, and the availability of important dependencies such as OpenMM, MDAnalysis, RDKit, ProLIF, PLIP, Flask, and NGLView.
For CI or scripted checks, use:
openmmdl check --strict
With --strict, the command exits with a non-zero return code if a required
dependency is missing.
Installation via Docker
OpenMMDL is mainly supported for Linux distribution systems, thus for Windows and MacOS the installation with docker may be preferred, due to docker creating an image with ubuntu:22.04.
For this first clone this repository:
cd ~
git clone https://github.com/wolberlab/OpenMMDL
change the path to the OpenMMDL folder and build the docker image from there:
cd OpenMMDL
docker build -t openmmdl_env .
This will build the OpenMMDL image with docker. Now that it is build you can access it through an interactive terminal:
docker run -it --name openmmdl_test openmmdl_env
From there you can access all the OpenMMDL entry points. Currently due to OpenMMDL Setup using flask it can be difficult to access it through the docker image.