Chemicals are used everywhere, with people and the environment being exposed to potentially harmful substances. The identification of intrinsic properties of chemical substances is needed to protect human health and the environment, however many of these chemicals transform either in the environment or in our bodies, while defining chemicals with an unknown or variable composition, complex reaction products or biological materials (so-called UVCBs) are increasingly important. The composition of these UVCB chemicals (estimated to be 25-40 % of high use chemical registries) can be variable or difficult to define, which means that the exchange of data about these chemical substances between resources is fraught with difficulty due to a lack of standards and compatible approaches, confounding FAIRification of these entities. Difficulties arise in: understanding the potential toxicological effects of such materials; linking to known biochemical transformations (e.g., found in Rhea); performing and interpreting metabolomics/exposomics experiments; data dissemination; data reuse; data workflows; and in formulating regulatory actions (e.g., under REACH). Several large projects funded by the European Union (e.g. CLP, NanoCommons, ZeroPM), as well as locally funded projects (e.g. VHP4Safety) try to address these issues, but desperately need these FAIR issues solved using cheminformatics approaches.
The aim of this project is to gather a critical mass of key cheminformatics resources and representatives to work on tackling the challenges with UVCBs and chemical transformations/metabolism. Our project builds on several previous BioHackathon efforts: Better integration with knowledge from Wikidata (Project 32 BioHackathon 2021) Include super and substructure searching (Elixir Czech Republic Service IDSM) FAIR Identifier Mapping, leading to PubChemLite (Project 27 BioHackathon 2019) Create metabolic and adverse outcome pathway models, which link better to other databases (Project 32 BioHackathon 2021) Extension and Continuous Integration of Cheminformatics Resources and Applications (Project 13, BioHackathon 2020)
- daily notes on HackMD.io, regular RO archive: https://gist.github.com/egonw/d10c0904387fcf8fab6ca5b0ce1c27f9
- BioHackRxiv reporting, draft: https://github.com/egonw/biohackeu22-project26
Bioschemas Data Platform Interoperability Platform Metabolomics Tools Platform
Project Number: 26
- Evan Bolton: GitHub ORCID
- Emma Schymanski: GitHub ORCID
- Anne Morgat: GitHub
- Egon Willighagen: GitHub ORCID
- Anjana Elapavalore: GitHub ORCID
- Dominik Martinát: GitHub ORCID
Improved handling of transformation product pairs and reactions (Rhea/PubChem/NORMAN-SLE/WikiPathways) Cheminformatics approaches for dealing with UVCBs (MInChI, “Concepts”) including first applications and examples Automated workflows for FAIR data deposition to PubChem, Rhea, WikiPathways and other key resources
Knowledge of chemistry, biochemistry, and/or toxicology is wanted. Our tools work in various programming languages (Java, Groovy, Python, R), as well as having different ways to integrate data (API, SPARQL, graph databases). Last, data curation is of importance to us (which does not always require programming skills), as well as being able to link to other databases, so we also welcome participants who have knowledge in this area.
Number of expected hacking days: 4