Large-Scale Bioactivity Data of Traditional Chinese Medicine

These files work with KNIME diagram to handle all training data and external validation data. 1. TCM_inchikey.xlsx includes all chemical structure data from 16 TCM databases. 2. After data augmentation, the original TCM structure data was expanded and stored as the 2000w_inchikey.txt. 3. Then TCM_inchikey.xlsx and 2000w_inchikey.txt were separatly mapped on the ChEMBL32 to acquire the bioassay data. Due to ChEMBL32 was acquired by SQL, please check KNIME diagram stored in github site for ChEMBL32 data and the mapped bioassay data. 4. After the screening process strored in KNIME, the training_data.zip was acquired. 5. BindingDB_bioassay.csv and pubchem_bioassay.csv were used to form the external validation data, that is external data.zip.