Deep-Learning-Driven Discovery of SN3-1, a Potent NLRP3 Inhibitor with Therapeutic Potential for Inflammatory Diseases

The official implementation of the algorithm used in 'Deep-Learning-Driven Discovery of SN3-1, a Potent NLRP3 Inhibitor with Therapeutic Potential for Inflammatory Diseases' (under review). This is our first public release version, so feel free to open an issue if you meet trouble applying our code.

Environments

Install via Conda

# Clone the environment
conda env create -f environment.yaml
#install rascorer
cd rascorer
git clone https://github.com/reymond-group/RAscore.git
pip install --editable .
# Activate the environment
conda activate deepoptc

Molecule Generation

We provide the checkpoint within the ckpt folder, download first to run the generation code.

To run the generation, first write the spliting smiles as the following form, test.smi

O=C([*])NC1=C2CCCC2=CC=C1C3=C(Cl)C(Cl)=CC=C3  O=S(C1=CC(C(O)(C)C)=CS1)(N*)=O  O=S(C1=CC(C(C)(O)C)=CS1)(NC(NC2=C(CCC3)C3=CC=C2C4=C(Cl)C(Cl)=CC=C4)=O)=O

Then modify the configuration file examples/decorate_scaffolds.json to include the input file and output file.

Generate attachment fragments by running:

python input.py examples/decorate_scaffolds.json

Cluster and Score

To cluster the generated fragments and score them according to Synthesizability, run the following:

python cluster_score/clustering.py output.csv RA_path output.sdf

This Code is build upon Lib-INVENT and RAscore. We would like to acknowledge their contribution.