Harmonized Drug Dataset (HDD)

Harmonized Drug Dataset (abbreviated as HDD) integrates comprehensive metadata for small‑molecule compounds, combining structural, physicochemical, regulatory, toxicological, and experimental assay information from PubChem, LINCS, Cell Painting, DeepChem resources, and multiple toxicology databases. It includes curated binding affinities, bioassay outcomes, adverse drug reactions, toxicity measurements, and machine‑learning‑ready molecular fingerprints. The dataset is designed to support large‑scale modeling, compound characterization, and toxicology prediction workflows. Version 2.3 (HDD_v2.3) is updated with ATC codes for 3202 compounds. The Anatomical Therapeutic Chemical (ATC) Classification System, maintained by the WHO, organizes active pharmaceutical ingredients according to the organ or system they act on, as well as their therapeutic, pharmacological, and chemical properties. Each drug is assigned a unique code across five hierarchical levels, moving from broad anatomical group down to the specific chemical substance.