Open science runs on a patchwork of repositories, standards, and tools that rarely fit together cleanly. Easing that fragmentation has been one of CEDAR’s central goals — letting a research community define its metadata standards once and then use them anywhere. A new post from the Dryad team, Dryad and CEDAR: Supporting more research communities with generalist infrastructure, shows this idea working in the wild.

The example is one we find especially striking: metadata templates we originally developed for the NIH Human BioMolecular Atlas Program (HuBMAP) are now being used to describe datasets in Dryad — an entirely different, general-purpose repository — without any new software, or even a new template, being written. The Dryad and OSF integrations we announced back in 2024 are now delivering that portability in practice.

How it works: a CEDAR template defines the structure of a discipline’s metadata — its fields, its controlled terms, and its value constraints. Turning that specification into a form a researcher can fill in is the job of the CEDAR Embeddable Editor (CEE). It is a lightweight Web Component that a platform embeds just once. From any template, it renders the matching entry form on the fly and returns the researcher’s answers as structured JSON-LD. Because the CEE builds its forms directly from templates, a platform can add or update a standard without writing new interface code. In Dryad, the CEE layers these discipline-specific templates on top of the repository’s default DataCite FAIR metadata. As a dataset is submitted, Dryad scans its title and abstract to suggest relevant templates, which the researcher completes in the embedded editor. The finished metadata is stored with the dataset as JSON and can be viewed in a read-only CEE on the dataset’s landing page.

Built for HuBMAP, reused in Dryad for free

As we described in Scientific Data, HuBMAP’s Data Coordination Working Group had developed several dozen reporting guidelines — one per assay type (RNAseq, FACS, histology, ATACseq, and more) — which the CEDAR team encoded as machine-actionable templates using the CEDAR Template Designer. Those same templates now help structure the metadata behind the HuBMAP Data Portal.

Because CEDAR templates are portable by design, Dryad was able to adopt more than 35 of these HuBMAP templates through the CEE at essentially no additional cost — no new modeling, no bespoke integration for each standard. Dryad reports that more than 65 datasets have already been enriched with this HuBMAP metadata. Templates authored once, for one community, now serve another “for free.”

This is the principle we care about most: research communities can capture rich, semantically precise metadata without each having to build and maintain its own specialist infrastructure. As Dryad puts it, the result is a system that is “both scalable and specific.”

Read Dryad’s full post, Dryad and CEDAR: Supporting more research communities with generalist infrastructure.