A workflow that runs fred or diffdock to screen small molecules for receptor proteins. The program can be run as a local script or a colmena workflow.

We will set up the run environment with Anaconda/Miniconda.

The base environment can be setup with the env.yml file in the repo.

git clone https://github.com/hengma1001/bvbrc_docking.git
cd bvbrc_docking
conda env create -n bvbrc_docking -f env.yml
pip install .

The required torch modules are installed using pip. It's recommand the user download the correct CUDA version for pytorch. (Note: test the pyg with conda install to reduce compilation time. )

pip install torch==2.1.0 torchvision==0.16.0 torchaudio==2.1.0 
pip install torch-scatter==2.1.2 torch-sparse==0.6.18 torch-cluster==1.6.3 torch-geometric==2.4.0

Next we need to setup DiffDock, along with esm. The Diffdock uses the esm package to obtain protein embeddings and structures.

git clone https://github.com/gcorso/DiffDock.git
cd DiffDock
git checkout a6c527536d286a693d1d6f9fc97142506ef76179
git clone https://github.com/facebookresearch/esm
cd esm
git checkout ca8a710

python -m bvbrc_docking.run_local -c runs/fred_local.yml