Download
Download the complete THEOBROMA dataset. Each record includes license_tier and all_sources columns for provenance tracking.
Programmatic access (REST API)
THEOBROMA provides a JSON API for programmatic access. All endpoints return JSON by default.
GET /api/search | Search compounds by name, SMILES, InChIKey, organism, kingdom, region, or source. Params: q, type, limit, offset, format (json/csv) |
|---|---|
GET /api/compound/<id> | Full details for a single compound by comp_id |
GET /api/autocomplete | Name autocomplete. Params: q (min 2 chars) |
GET /api/organisms | Organism name autocomplete. Params: q |
GET /api/stats | Database summary statistics |
GET /api/compound/<id>/license-provenance | Per-source license attestation chain (audit trail) for a compound |
GET /api/stereoisomers/<id> | All stereoisomers of a compound (14-character InChIKey match) |
GET /api/similarity | Structural similarity search. Params: smiles, metric (morgan/maccs/chemberta/nafm) |
GET /api/filter_options | Available values per filter type (kingdoms, sources, regions, classes, taxonomic ranks) |
GET /api/bulk | Streaming bulk export (no row limit). Params: cols, format, license |
Examples
curl "https://theobroma.l3s.uni-hannover.de/api/search?q=curcumin&type=name&limit=10" curl "https://theobroma.l3s.uni-hannover.de/api/search?q=Streptomyces&type=organism&format=csv" curl "https://theobroma.l3s.uni-hannover.de/api/compound/THEO_0856795" curl "https://theobroma.l3s.uni-hannover.de/api/stats"
HuggingFace dataset
The full dataset and precomputed vectors are available on HuggingFace.