Database of herbal metabolites

The Herbal Metabolites Database (HerbalMDB) integrates comprehensive information on secondary metabolites, including basic details, similarities, taxonomic data, ADMET properties, and perturbation profiles. Data quality was ensured through deduplication and normalization using RDKit. The dataset was enriched by generating Morgan fingerprints and assessing similarities via Tanimoto coefficients, classifying metabolites with NPClassifier, predicting ADMET properties with ADMET-AI, and determining gene expression profiles using DeepICER. These efforts enhance the dataset's comprehensiveness and utility.