ga4gh-discovery / data-connect

Standard for describing and searching biomedical data developed by the Global Alliance for Genomics & Health.
Apache License 2.0
24 stars 14 forks source link

Consider a specific schema field for unit #105

Open ianfore opened 3 years ago

ianfore commented 3 years ago

Background In Search discussion on 2 March the following scope was discussed - see document What is the scope of Search?

  1. agree on how we search for things with a given data model
  2. agree on how we describe data models
  3. agree on the data models themselves

The conclusion was that 1 and 2 are in scope. 3 is beyond scope for specific models but is facilitated by 2. Notably models defined in 3 would be recorded via 2.

Issue The issue here concerns 2. The schema provided by /tables//info are how the models in 2 are made available. The specific request is that the metaschema used for 2 should provide explicit handling of unit of measure in structured form. The current metaschema is JSON-Schema. In Search examples to date unit have been embedded in $comment. e.g.

{'age':{
   "type": "number",
   "$comment": "UNIT 'Years'",
   "description": "Participant reference age"
}

Finding some way to include unit should be considered. Likely using an existing semantic standard for unit of measure. The following is a conceptual illustration of a minimal change to accomplish this.

{'age':{
   "type": "number",
   "unit": "Years",
   "description": "Participant reference age"
}

As a test case, the handling of unit should allow for automated use of the kind of age mappings illustrated in this FASP Script.