Dear Vidya,
Thank you for your interest in using COCONUT database for virtual screening.
A quick clarification first, the "Approved" filter in the search interface refers to curation status, not regulatory/clinical approval. Every molecule in COCONUT carries a status (approved, draft, revoked, etc.) that reflects whether it has passed our curation and structure-validation pipeline. "Approved" is therefore the active, curated dataset — it is not a list of approved drugs.
With that clarified please find the details below
• A bulk downloadable file (preferably SDF/SMILES) containing only the compounds classified as approved; or
The SDF and CSV files on our download page (https://coconut.naturalproducts.net/download) already contain only the active/approved molecules. No separate file is needed:
• coconut_sdf_2d_lite-09-2026.zip / coconut_sdf_2d-09-2026.zip — 2D SDF (lite = structures + core identifiers; full = with annotations)
• coconut_sdf_3d-09-2026.zip — 3D SDF with RDKit-generated coordinates (not minimised), which may be a useful starting point for docking
• coconut_csv_lite-09-2026.zip / coconut_csv-09-2026.zip — CSV, including canonical SMILES
Only the full PostgreSQL dump (coconut-dump-09-2026.sql) includes inactive/revoked entries, so avoid that one unless you want the complete history.
An API/query method through which the approved compounds can be retrieved in bulk.
The REST API (https://coconut.naturalproducts.net/api-documentation) supports paginated searches, and the search endpoint can be filtered by status. For a full-database pull, however, the bulk files above are far more efficient and are what we recommend.
All COCONUT data is released under CC0, so there are no restrictions on its use. Good luck with the screening — we'd be glad to hear how it goes.
Regards,
Chandu
—
Venkata Chandra Sekhar Nainala
Scientific Lead Architect and Developer
Friedrich-Schiller-University Jena
Faculty of Chemistry and Geosciences
Institute of Inorganic and Analytical Chemistry
Lessingstraße 8
07743 Jena
On 22 Sep 2026, at 06:30, V S Vidya vsvidya@jssuni.edu.in wrote:
Dear COCONUT Team,
I am a PhD researcher working in the area of structure-based drug discovery. I am planning to perform a virtual screening study using natural products available in the COCONUT database.
While browsing the COCONUT compound database, I noticed that the compounds can be filtered using the “Approved” option. I would like to download the complete set of compounds corresponding to this approved category for use in a virtual screening workflow.
Since the number of compounds is very large, I would like to know whether COCONUT provides:
• A bulk downloadable file (preferably SDF/SMILES) containing only the compounds classified as approved; or
• An API/query method through which the approved compounds can be retrieved in bulk.
If a separate approved-compound dataset is not available, could you please advise on the recommended method for retrieving this subset from COCONUT?
The downloaded compounds will be used for academic research and structure-based virtual screening.
Thank you for your assistance.
Kind regards,
V S Vidya
PhD Scholar
Department of Biotechnology & Bioinformatics
JSS Academy of Higher Education & Research
Mysuru, India
Coconut-support mailing list -- coconut-support@lists.nfdi.de
To unsubscribe send an email to coconut-support-leave@lists.nfdi.de
Dear Vidya,
Thank you for your interest in using COCONUT database for virtual screening.
A quick clarification first, the "Approved" filter in the search interface refers to curation status, not regulatory/clinical approval. Every molecule in COCONUT carries a status (approved, draft, revoked, etc.) that reflects whether it has passed our curation and structure-validation pipeline. "Approved" is therefore the active, curated dataset — it is not a list of approved drugs.
With that clarified please find the details below
The SDF and CSV files on our download page (https://coconut.naturalproducts.net/download) already contain only the active/approved molecules. No separate file is needed:
• coconut_sdf_2d_lite-09-2026.zip / coconut_sdf_2d-09-2026.zip — 2D SDF (lite = structures + core identifiers; full = with annotations)
• coconut_sdf_3d-09-2026.zip — 3D SDF with RDKit-generated coordinates (not minimised), which may be a useful starting point for docking
• coconut_csv_lite-09-2026.zip / coconut_csv-09-2026.zip — CSV, including canonical SMILES
Only the full PostgreSQL dump (coconut-dump-09-2026.sql) includes inactive/revoked entries, so avoid that one unless you want the complete history.
The REST API (https://coconut.naturalproducts.net/api-documentation) supports paginated searches, and the search endpoint can be filtered by status. For a full-database pull, however, the bulk files above are far more efficient and are what we recommend.
All COCONUT data is released under CC0, so there are no restrictions on its use. Good luck with the screening — we'd be glad to hear how it goes.
Regards,
Chandu
—
Venkata Chandra Sekhar Nainala
Scientific Lead Architect and Developer
Friedrich-Schiller-University Jena
Faculty of Chemistry and Geosciences
Institute of Inorganic and Analytical Chemistry
Lessingstraße 8
07743 Jena