Source Code of 《Multi-modal data fusion for supervised learning-based identification of USP7 inhibitors: a systematic comparison》

• Python >= 3.8

• numpy

• pandas

• scikit_learn

• pytorch

• tqdm

│ LICENSE
│ README.md
│ requirements.txt     # Recommended Environment
├─data     # Include the USP7 dataset, decoy dataset and SMILES enumeration dataset
├─results     # Stored the results of all experiments, and Tables in Additional file 1
└─src     # Stored the main codes (cutoff values = 0.5,1,10, SMOTE, decoy selection, SMILES enumeration)
      ├─DL-1      # The source code of experiment 14 in Group VI-X.
                         # Experiment 8-13 only needs to change the inputs.
      ├─DL-2      # The source code of Experiment 15 in Group XI-XV.
      └─ML         # The source code of experiment 7 in Group I-V.
                         # Experiment 1-6 only needs to change the inputs.

• The recommended hyperparameters for each ML or DL models are available in the paper`s Additional file 1.
• We provide the best models and prediction scripts of Exp7 in Group I, Exp15 in Group XI and Exp14 in Group VI. If you need the prediction script of other experiments, you can simply replace and modify it.

Hovi Tang: hovi@shu.edu.cn