The Redevelopment of ChEBI

We are delighted to announce that ChEBI’s beta version is now out and can be accessed at: https://www.ebi.ac.uk/chebi/beta/

As many of you know, ChEBI is a widely used database and ontology of chemical entities of biological interest. It has a growing user base numbered in the hundreds of thousands and with millions of accesses each year. ChEBI has for the past >20 years provided the biological community with access to definitive small molecule information and data, including accurate representations of often complex chemical structures. Additionally, ChEBI is used widely as a reference database and ontology by many important bioinformatics resources such as Rhea, MetaboLights, UniProt, GO, IEDB, Reactome, amongst others. ChEBI is, thus, a fundamental and core component of the global bioscience infrastructure.

Outdated infrastructure 

Back in 2021, ChEBI's software infrastructure was increasingly fragile. ChEBI's current software code base dates back to its creation in the early 2000s and was ever-more difficult and costly to maintain, with significant amounts of redundant and non-functioning code. Put simply, if we did not act to undertake the important task of re-engineering and re-writing ChEBI, it would have been increasingly difficult to keep the resource running, and very likely that there would have come a point where it could have been turned off due to a fundamental incompatibility with modern computational infrastructure.

What have we delivered so far…. 

The "new ChEBI" was written using modern software technologies, and therefore is easier to support, more flexible, more efficient and overall, more sustainable. Over the past four years, we undertook a number of individual tasks that collectively delivered on this overall goal:

1. Migrated the ChEBI database to a new streamlined schema within the PostgreSQL system.
2. Developed a set of efficient web services using the REST (REpresentational State Transfer) protocol to replace the current deprecated SOAP-based system. 
3. Developed methods for exporting the data in ChEBI in a variety of file formats. Find out more about ChEBI 2.0 data products here
4. Created a new set of searching capabilities based on new technologies such as elasticsearch, with chemical structure storage and searching based on the widely used open-source RDKit system.
5. Developed a new and enhanced public web interface using modern technologies including reusable web components. 
6. Provided a widget to enable users to visualise and manipulate the ChEBI ontology which is a unique component of ChEBI.
7. Organised a workshop for key resources that depend upon ChEBI, to update on progress of the project and receive feedback.

What’s next?

In the next few months, our focus will be to develop a new curation/submissions tool that enables ChEBI curators and external users to efficiently annotate and add data into ChEBI, ensure that documentation is updated, and finally turn off the old interface.