CYP2A6 Inhibition Prediction Training Dataset

A curated dataset constructed using compounds retrieved from natural products databases. Molecular docking simulations were performed using rDock against the CYP2A6 protein, yielding docking scores that were used as the target endpoint for modelling. For each compound, a comprehensive set of molecular descriptors was calculated using the RDKit library. This included 217 2D descriptors derived from molecular topology and physicochemical properties, as well as 7 3D descriptors capturing spatial and conformational characteristics. The final dataset integrates docking-derived bioactivity scores with ligand-based molecular representations, enabling the development of machine learning models for predicting compound–protein interactions.