Forked from the official repo and upgraded the python and torch version for Amper GPU.

Following are the original README:

---

Codes for "MONN: a Multi-Objective Neural Network for Predicting Pairwise Non-Covalent Interactions and Binding Affinities between Compounds and Proteins"

The benchmark dataset described in this paper can be found in ./data/, and the creation of this dataset can be reproduced by the protocol in ./create_dataset/.

Before running the MONN model in ./src/, please first use ./src/preprocessing_and_clustering.py to produce necessary files.

For cross validation, e.g., using IC50 data, new-compound setting and clustering threshold 0.3, run:

python CPI_train.py IC50 new_compound 0.3

Requirements:

Python2.7

rdkit (for preprocessing)

Pytorch >= 0.4.0

scikit-learn

License

This software is copyrighted by Machine Learning and Computational Biology Group @ Tsinghua University.

The algorithm and data can be used only for NON COMMERCIAL purposes.

---

Upgrade Notes

Preset conda environment

git clone https://github.com/JinyuanSun/MONN.git
cd MONN
conda env create -f environment.yml
conda activate monn

More details about step 3:

Step3: calculate the non-covalent interactions between proteins and ligands
Extract the non-covalent interactions by using PLIP (https://github.com/ssalentin/plip/) Put the result files in ./plip_result/ Note: command for using PLIP:

python plipcmd.py -f xxx.pdb -t --name xxx_output

The plip is a good package, relies on openbable, I recommand using docker:

docker pull pharmai/plip:latest
docker run --rm -v ${PWD}:/results -w /results -u $(id -u ${USER}):$(id -g ${USER}) pharmai/plip:latest -f 1o41.pdb -t --name output_1o41

Also you are welcomed to build it from source code. For some reason, as far as I konw, the conda and pip installation just won't work.
With docker:

ls ????.pdb|awk -F "." '{print"docker run --rm -v ${PWD}:/results -w /results -u $(id -u ${USER}):$(id -g ${USER}) pharmai/plip:latest -f "$1".pdb -t --name ../plip_result/output_"$1""}' > job_list.sh
parallel --jobs $num_of_threads < job_list.sh

If you managed to install the plipcmd.py, run the following in ./pdb_files/:

ls ????.pdb|awk -F "." '{print"plipcmd.py -f "$1".pdb -t --name ../plip_result/output_"$1""}' > job_list.sh
parallel --jobs $num_of_threads < job_list.sh

Good luck with it!

Citation

@article{LI2020308,
title = {MONN: A Multi-objective Neural Network for Predicting Compound-Protein Interactions and Affinities},
author = {Shuya Li and Fangping Wan and Hantao Shu and Tao Jiang and Dan Zhao and Jianyang Zeng},
journal = {Cell Systems},
volume = {10},
number = {4},
pages = {308-322.e11},
year = {2020},
issn = {2405-4712},
doi = {https://doi.org/10.1016/j.cels.2020.03.002},
url = {https://www.sciencedirect.com/science/article/pii/S2405471220300818}
}

Warning

No guarantee of 100% reproduction results in the paper after modifications, if that is what you want, check the official repo!!!