MetalProGNet: A Structure-based Deep Graph Model Specific For Metalloprotein-Ligand Interaction Predictions; MetalProGNet was developed based on our previous IGN framework.
- Step1: Clone the Repository
git clone https://github.com/zjujdj/MetalProGNet.git- Step2: Download the trained models and conda-packed env
download url: dgl430_py37_gpu.tar.gz; model_save.zip
cd MetalProGNet && unzip model_save.zip && tar -xzvf dgl430_py37_gpu.tar.gz -C /home/conda_env/dgl430_py37_gpu
source activate /home/conda_env/dgl430_py37_gpu
conda-unpack- Step3: Give executable privileges to the PLANTS program
chmod +x ../plants/PLANTS1.2_64bit ../plants/SPORES_64bit- Step4: Using plants_protein_prep.py for protein preparation
python3 plants_protein_prep.py --num_process=4 --receptor_path=../receptors/ --plants_path=../plants/- Step5: Using plants_ligand_prep.py for ligand preparation
python3 plants_ligand_prep.py --num_process=10 --ligand_file_path=../ligand_file/ --plants_path=../plants/ --temp_path=../temp/ --dst_path=../prepared_ligands/- Step6: Using plants_docking.py for molecular docking
cd .. && rm -rf docking_runing && mkdir -p docking_running
cp ./scripts/plants_docking.py docking_running
cd docking_running
python3 plants_docking.py --num_process=10 --top1_pose_path=../top1_pose/ --top1_pose_sdf_path=../top1_pose_sdf/ --receptor_path=../receptors/ --plants_path=../plants/ --dst_path=../prepared_ligands/ --config_file_path=../config_file/ --docking_running_path=../docking_running/
cd .. && rm -rf docking_running- Step7: Using predict_mt_chembl.py for binding affinity prediction
cd scripts
python3 predict_mt_chembl.py --num_process=10 --bin_size=5 --batch_size=512 --sdfs_path=../top1_pose_sdf/ --protein_path=../receptors/ --temp_path=../temp/ --dock_engine=plants --csv_path=../csv_files/ --work_name=test
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent