A5093671268- Machine Learning in Bioinformatics
- Genetics, Bioinformatics, and Biomedical Research
- Scientific Computing and Data Management
- Research Data Management Practices
- Biomedical and Engineering Education
ZB MED - Information Centre for Life Sciences
2024
University of Bonn
2023
Here we present a proof-of-concept using Bioschemas, RO-Crate and Signposting as lightweight approach to FDOs describing research outcomes exposed on GitHub pages. Research artifacts produced by development team, namely the SemTec team at ZB MED, are via pages enriched with structured metadata schema.org Bioschemas. corresponding same project put together in an RO-Crate. is used landing The combination of these three elements facilitates rich FAIR for artifacts.
To advance the use of machine learning to address humanity’s grand challenges such as understanding disease conditions and biodiversity loss in anthropocene, it is important promote FAIR AI-ready datasets, since data scientists bioinformaticians spend 80% their time finding preparation. Metadata descriptors for datasets are pivotal creation models they facilitate definition strategies discovery, feature selection, cleaning, pre-processing. ML-ready whether by design or after pre-processing,...
As part of the BioHackathon Europe 2023, we here report from progress hackathon project #15: "Enabling FAIR Digital Objects with RO-Crate, Signposting and Bioschemas". We added to three existing resources, made a Chrome browser extension show headers. RO-Crate two explored making hybrid FDO using both Handle PID Record Signposting/RO-Crate approach.