Metadata standards for animal genomics
DCAT-based dataset descriptions in JSON-LD, with controlled vocabularies, ontology mappings, OCA schemas and automated validation, so public genotyping datasets are curated once and reused across projects.
A dataset is only reusable if someone outside the project that made it can find it and understand it. That's what the F.A.I.R. principles ask for (findable, accessible, interoperable, reusable), and at Genovalia, that's my job.
The model
I designed Genovalia's metadata model on DCAT (the W3C standard for describing datasets in catalogues), written as JSON-LD (JSON where each field points to a shared definition), with controlled vocabularies and mappings to established ontologies. The structure of each dataset's data is described with OCA (Overlays Capture Architecture) schemas. The catalogue and the API store, index and serve that model directly, with no conversion in between.
Curation pipelines
Much of the useful data already exists in public repositories such as Dryad, Zenodo and Borealis. The pipelines I built prepare those genotyping datasets for the catalogue:
- describe them in DCAT,
- attach their OCA schemas,
- validate everything automatically and produce a quality report,
- ingest the result into the catalogue.
Once a dataset is curated this way, the next project can find it and reuse it instead of starting over.