COCONUT 2.0: The Collection of Open Natural Products Database
Abstract
Natural Products (NPs) are bioactive small molecules synthesized by living organisms, long recognized for their pharmaceutical and industrial potential. Despite numerous databases, a comprehensive online resource compiling digitally documented NP structures in one place was lacking. The COlleCtion of Open Natural prodUcTs (COCONUT) is a comprehensive open database that aggregates data from numerous public sources. Originally launched in 2021 [ 1 ], COCONUT has undergone a major overhaul with version 2.0 to improve its infrastructure, data quality, and user experience. COCONUT 2.0 [ 2 ] contains over 695,000 unique natural product structures compiled from 63 data collections. COCONUT 2.0 provides core information including chemical structures, names/synonyms, source organisms, geographic origins, and literature references. Computed molecular properties and chemical classifications are also included. Key improvements in version 2.0 include a complete redesign of the web interface and backend architecture, enhanced data curation and standardisation processes, new features for community-driven curation and data submission, improved search capabilities—including structure, substructure, and similarity searches—a REST API for programmatic data access, downloadable datasets in multiple formats, and detailed provenance tracking and audit logs. Since COCONUT is an aggregate, issues from its constituent datasets, such as occurrences of synthetic compounds and inaccurate metadata,inevitably appear in the database. By harnessing advanced language models, agentic flows, and community contributions, the team is systematically refining the dataset to eliminate misclassifications and expand genuine natural product entries. With its open licenses, COCONUT remains a robust, evolving resource dedicated to advancing natural product research and its diverse applications. Publication History Article published online: 05 March 2026 © 2026. Thieme. All rights reserved. Georg Thieme Verlag KG Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany
Fields of Study & Subjects
Research Data Management Practices
Computational Drug Discovery Methods
Bioactive natural compounds
Publication details
- Journal / source
- Planta Medica
- Volume
- 92
- Publication date
- 1 March 2026
- Language
- EN