Skip to content

sssom-rosetta

Dutch healthcare data uses many different models. OMOP CDM, FHIR, SNOMED CT and LOINC are international standards. ONZ-G, Z-Index and the DHD thesauri are Dutch. Today, the links between them live in spreadsheets that go out of date fast.

sssom-rosetta brings these models together in one place. It checks that every mapping points to something real. It uses Turtle (Terse RDF Triple Language) as the shared format for ontologies and vocabularies, while SSSOM is used for the for mappings, so they stay easy to read, review and reuse.

Two kinds of building blocks

  • Ontologies are information models. We map them to each other, one pair at a time, by hand. See Mappings for the current mapping set.
  • Vocabularies are large code lists and terminologies, such as OMOP's standard vocabularies and the DHD thesauri. We combine them into one graph using maplib, so every code becomes something other mappings can point at.

Where this is going

The aim of sssom-rosetta is to expose these vocabularies programmatically — concepts, hierarchy, relationships, mappings, and search over a versioned, snapshot-per-release store — so the integrated vocabulary can drive both mapping authoring and downstream ETL. We take inspiration from (omophub-python).

See the roadmap for what's planned next.