End-to-end protein-ligand complex structure generation with diffusion-based generative models

Setup Environment

Clone this repository and install dependencies:

git clone https://github.com/shuyana/DiffusionProteinLigand.git
cd DiffusionProteinLigand
conda env create -f environment.yml
conda activate dpl

Download model parameters:

gdown --fuzzy --folder https://drive.google.com/drive/u/1/folders/1AAJ4P5EmQtwle9_eSeNMcF-KMWObksxZ

Additionally, TMalign is required to align generated structures. You can install it as follows:

wget https://zhanggroup.org/TM-align/TMalign.cpp
g++ -static -O3 -ffast-math -lm -o TMalign TMalign.cpp
chmod +x TMalign
export PATH="/path/to/TMalign:$PATH"

Sample generation

Generate complex structures with the protein structure-free model (DPL):

python generate.py \
    --ckpt_path "checkpoints/DPL_v1.ckpt" \
    --output_dir "workdir/generate/example_DPL" \
    --protein "LSEQLKHCNGILKELLSKKHAAYAWPFYKPVDASALGLHDYHDIIKHPMDLSTVKRKMENRDYRDAQEFAADVRLMFSNCYKYNPPDHDVVAMARKLQDVFEFRYAKMPD" \
    --ligand "Cc1ccc2c(c1c3cc(cc4c3nc([nH]4)C5CC5)c6c(noc6C)C)cccn2" \
    --num_samples 8

Alternatively, the protein structure-dependent model (DPL+S) can be used:

wget https://files.rcsb.org/download/6MOA.pdb
python generate.py \
    --ckpt_path "checkpoints/DPLS_v1.ckpt" \
    --output_dir "workdir/generate/example_DPLS" \
    --protein "6MOA.pdb" \
    --ligand "Cc1ccc2c(c1c3cc(cc4c3nc([nH]4)C5CC5)c6c(noc6C)C)cccn2" \
    --num_samples 8

Note that an input protein structure must be given as a PDB file in this case.

Besides, you can specify a reference protein structure to be used for the alignment of results:

python generate.py \
    --ckpt_path "checkpoints/DPL_v1.ckpt" \
    --output_dir "workdir/generate/example_DPL_ref" \
    --protein "LSEQLKHCNGILKELLSKKHAAYAWPFYKPVDASALGLHDYHDIIKHPMDLSTVKRKMENRDYRDAQEFAADVRLMFSNCYKYNPPDHDVVAMARKLQDVFEFRYAKMPD" \
    --ligand "Cc1ccc2c(c1c3cc(cc4c3nc([nH]4)C5CC5)c6c(noc6C)C)cccn2" \
    --num_samples 8 \
    --ref_path "6MOA.pdb"

This is used only for alignment and does not affect the generation process itself.

The argument num_steps can be modified from the default of 64 to reduce execution time:

python generate.py \
    --ckpt_path "checkpoints/DPL_v1.ckpt" \
    --output_dir "workdir/generate/example_DPL_fast" \
    --protein "LSEQLKHCNGILKELLSKKHAAYAWPFYKPVDASALGLHDYHDIIKHPMDLSTVKRKMENRDYRDAQEFAADVRLMFSNCYKYNPPDHDVVAMARKLQDVFEFRYAKMPD" \
    --ligand "Cc1ccc2c(c1c3cc(cc4c3nc([nH]4)C5CC5)c6c(noc6C)C)cccn2" \
    --num_samples 8 \
    --ref_path "6MOA.pdb" \
    --num_steps 24

Training

Download the PDBbind dataset from https://zenodo.org/record/6408497 and unzip it.

Move the resulting PDBBind_processed directory to data/.

Preprocess the dataset:

python preprocess_pdbbind.py

Finally, run the training script:

python train.py \
    --num_workers 8 \
    --batch_size 1 \
    --accumulate_grad_batches 8 \
    --no_cb_distogram \
    --save_dir "workdir/train/example_DPL" \
    --single_dim 256 \
    --pair_dim 32 \
    --num_blocks 4

where the no_cb_distogram argument makes the model protein structure-free.

Please modify the batch_size and accumulate_grad_batches arguments according to your machine(s).

Default values can be used to reproduce the settings used in our paper:

python train.py \
    --num_workers 8 \
    --batch_size 3 \
    --accumulate_grad_batches 8 \
    --no_cb_distogram \
    --save_dir "workdir/train/reproduce_DPL"

Citation

@article{Nakata2023,
    doi = {10.1186/s12859-023-05354-5},
    url = {https://doi.org/10.1186/s12859-023-05354-5},
    year = {2023},
    month = jun,
    publisher = {Springer Science and Business Media {LLC}},
    volume = {24},
    number = {1},
    author = {Shuya Nakata and Yoshiharu Mori and Shigenori Tanaka},
    title = {End-to-end protein{\textendash}ligand complex structure generation with diffusion-based generative models},
    journal = {{BMC} Bioinformatics}
}

Acknowledgements