Hacking for Chem (again!)

German and European Hackathons

Two hackathons with NFDI4Chem participation took place during the last months. 

The first one was the 2023 ELIXIR BioHackathon Europe. It was organised in hybrid form on the wonderful Campus Belloch near Barcelona and online. 

Several NFDI4Chem’ers were part of the project to create knowledge graphs from our data ressources MassBank, Chemotion Repository and nmrXiv. The main activities were to import the SchemaOrg metadata from our ressources MassBank, Chemotion and nmrXiv into (Knowledge), as well as ressources from the DataPlant consortium into Graph databases, and then linking and querying across these data sources. 

The next hackathon was the 2nd BioHackathon Germany #BHG2023. Here, Oliver Koepler and Steffen Neumann organised a project on Ontologies and Terminology services ecosystem, since the world of ontologies and terminology services is a critical area for data interoperability in natural sciences and beyond.

Structuring the project across four pivotal areas

The Terminology Service (TS) Department: Dedicated to the efficient management and provision of terminologies, alongside developing innovative service components. Participants became familiar with how the TIB TS service (from TIB – Leibniz-Informationszentrum Technik und Naturwissenschaften und Universitätsbibliothek) works, and the information on how to get ontologies added to TIB TS was immediately put to practice by our neighboring NFDI consortium DataPlant, who now have a tailored DataPlant collection based on the joint landscape paper 10.3389/fpls.2023.1279694.

Consumer Department: Concentrating on the creation of user-friendly HTML widgets for term display and selection, and seamlessly integrating these tools into a range of research resources. 

Specifically, the ontology widgets were integrated into the Streamlit low-code web framework. The Google Spreadsheet add-on OntoMaton was updated to work with TIB TS (and the recent OLS4 update). The DataPlant team will connect their SWATE tool also to TIB TS.

We are very excited to also see the widgets integrated into Chemotion and LabIMotion to create customized forms that incorporate controlled terms from the TIB Terminology Service API. Our current challenge lies in managing the sheer volume of ontology and terminology terms, the goal is to devise solutions that efficiently recommend the most pertinent ontologies for specific term annotations within these annotation tools.

We were further able to collect many inspiring use cases for the upcoming #NFDI basic service project #TS4NFDI which will develop a central API Gateway to access Terminology Services of the NFDI. 

Additionally, we were exploring the broader application of ontology terms through the DefinedTerm concept in SchemaOrg, especially in the context of metadata collections shared across repositories and disciplines.

The unexpected (but highly welcome!) collaboration with the BioCypher project might allow us next year to query our knowledge graphs (remember the first Hackathon mentioned above ?) using natural language, where a large language model translates the request into a graph query. 

Oliver and Roman made some posts to LinkedIn and tooted on Mastodon.

The event was an exciting opportunity to collaborate and exchange ideas with a diverse group of ontology enthusiasts from various NFDI consortia and other innovative projects.

Finally, you are currently reading the first output from the Dissemination & Outreach Department, with more to come. We will contribute our experience to the NFDI Section Metadata. If things work out, you’ll see more in 2024 at NFDI’s own metadata hackathon or the EOSC event in October in Berlin. 

Please save the dates for two upcoming NFDI talks

1) Terminology services 4 NFDI on 22.01.2024 @ 16:00, and

2) SchemaOrg4NFDI on 18.03.2024 @ 16:00.