Draft of a schema for NeuroBagel's graph data format

[mih]

With this schema, I can convert the demo data shipped with NeuroBagel into YAML and back. I can do validation and all the other linkml goodness.

This is posted here, in order to fuel a data transformation pipeline from "our" schemas to NeuroBagel's in order to feed an instance with dataset information.

This schema draft is currently not rendered or hooked up to the validation and linting, because it does not meet our conventions (and will never). I need to allow for a schema-specific linter config first.

  warning  Slot has name 'schemaKey'  (standard_naming)
  warning  Slot has name 'hasContrastType'  (standard_naming)
  warning  Slot has name 'hasPipelineVersion'  (standard_naming)
  warning  Slot has name 'hasPipelineName'  (standard_naming)
  warning  Slot has name 'hasLabel'  (standard_naming)
  warning  Slot has name 'hasAge'  (standard_naming)
  warning  Slot has name 'hasSex'  (standard_naming)
  warning  Slot has name 'isSubjectGroup'  (standard_naming)
  warning  Slot has name 'hasDiagnosis'  (standard_naming)
  warning  Slot has name 'hasAssessment'  (standard_naming)
  warning  Slot has name 'hasFilePath'  (standard_naming)
  warning  Slot has name 'hasAcquisition'  (standard_naming)
  warning  Slot has name 'hasCompletedPipeline'  (standard_naming)
  warning  Slot has name 'hasSession'  (standard_naming)
  warning  Slot has name 'hasPortalURI'  (standard_naming)
  warning  Slot has name 'hasSamples'  (standard_naming)
  warning  Schema maps prefix 'ncit' to namespace 'http://ncicb.nci.nih.gov/xml/owl/EVS/Thesaurus.owl#' instead of using prefix 'NCIThesaurus'  (canonical_prefixes)
  warning  Schema maps prefix 'nidm' to namespace 'http://purl.org/nidash/nidm#' instead of namespace 'http://nidm.nidash.org/'  (canonical_prefixes)
  warning  Schema maps prefix 'np' to namespace 'https://github.com/nipoppy/pipeline-catalog/tree/main/processing/' instead of namespace 'http://www.nanopub.org/nschema#'  (canonical_prefixes)

[mih]

This is not useful as-is. The reason is that, in a generated JSON-LD context, the schema classes do not end up being defined as nb:Acquisition, for example, but come without the prefix. This makes NeuroBagel reject such documents.