Graphein depends on a number of other libraries for constructing graphs, meshes and adding features.

The source code for Graphein can be viewed at a-r-j/graphein


For full functionality, there are a number of additional (optional) installs for libraries not available via pip/conda. Please see below

Core Install#

At present, the simplest installation is via PyPI . The base install can also be performed using the provided conda environment.

pip install graphein         # For base install
pip install graphein[extras] # For additional featurisation dependencies
git clone
cd graphein
conda env create -f environment.yml
pip install .

Docker Install#

We provide two docker-compose files for CPU (docker-compose.cpu.yml) and GPU usage (docker-compose.yml) locally. For GPU usage please ensure that you have NVIDIA Container Toolkit installed. Ensure that you install the locally mounted volume after entering the container (pip install -e .). This will also setup the dev environment locally.

The Dockerfile is viewable here

docker-compose.cpu up -d --build    # start the container
docker-compose.cpu down             # stop the container
docker-compose up -d --build        # start the container
docker-compose down                 # stop the container

Dev Install#

The Dev install of Graphein contains additional dependendencies for development, such as testing frameworks and documentation tools. If you wish to contribute to Graphein, this is the installation method you should use.

Alternatively, if you wish to install Graphein in the dev environment (includes GPU builds (CUDA 11.1) of all relevant geometric deep learning libraries) you can use the provided conda environment:

pip install graphein[dev] # For dev dependencies
pip install graphein[all] # To get the lot
git clone
cd graphein
conda env create -f environment-dev.yml
pip install -e .  # Install in editable mode


We provide a devcontainer for the dev environment. This is a lightweight container that can be used to run the dev environment locally.

More information about devcontainers

Optional Dependencies#

However, there are a number of (optional) utilities DSSP, PyMol, GetContacts that are not available via PyPI:

conda install -c salilab dssp # Required for computing secondary structural features
conda install -c schrodinger pymol # Required for PyMol visualisations & mesh generation


Some of these packages have more involved setup depending on your requirements (i.e. CUDA). Please refer to the original packages for more detailed information

Installing Deep Learning Libraries#

Due to the many possible configurations of deep learning libraries, we deliberately do not provide a single install via PyPI. However, the conda dev environment described above contains GPU builds for CUDA 11.1 and PyTorch. The Dockerfile for the GPU build is provided in the docker-compose.yml file.

conda install -c pytorch pytorch
conda install -c pytorch3d pytorch3d #  NB requires fvcore and iopath
conda install -c dglteam dgl
conda install pytorch-geometric -c rusty1s -c conda-forge


GetContacts is an optional dependency for computing intramolecular contacts in .pdb files. We provide distance-based heuristics for this in graphein.protein.edges.distance so this is not a hard requirement.

Please see the GetContacts documentation for up-to-date installation instructions.

# Install dependencies
conda install scipy numpy scikit-learn matplotlib pandas cython seaborn
pip install ticc==0.1.4

# Install vmd-python dependencies
conda install netcdf4 numpy pandas seaborn  expat tk=8.5  # Alternatively use pip
brew install netcdf pyqt # Assumes is installed

# Set up vmd-python library
git clone
cd vmd-python
python build
python install
cd ..

# Set up getcontacts library
git clone
echo "export PATH=`pwd`/getcontacts:\$PATH" >> ~/.bash_profile
source ~/.bash_profile

# Test installation
cd getcontacts/example/5xnd --topology 5xnd_topology.pdb \
                        --trajectory 5xnd_trajectory.dcd \
                        --itypes hb \
                        --output 5xnd_hbonds.tsv
# Install dependencies
conda install scipy numpy scikit-learn matplotlib pandas cython
pip install ticc==0.1.4

# Set up vmd-python library
conda install -c vmd-python

# Set up getcontacts library
git clone
echo "export PATH=`pwd`/getcontacts:\$PATH" >> ~/.bashrc
source ~/.bashrc