Develop enums for Study, Biospecimen, Data File, Assay and add ontology terms

[lopierra]

Add/review/finalize enums/ontology terms for the following Study/Dataset fields (see notes here):

• Biospecimen Develop enumerations for the field: biospecimen.SampleType #169
• Assay Develop enumerations for the field: Assay #170
• Research Domain Develop enumerations for the field: Research Domain #174
• Participant Lifespan Stage Develop enumerations for the field: Participant Lifespan Stage #175
• Study Design Develop enumerations for the field: Study Design #179
• Clinical Data Source Type Develop enumerations for the field: Clinical Data Source Type #176
• Data Category Develop enumerations for the field: Data Category #177
• Data Type Develop enumerations for the field: Data Type #180
• Experimental Strategy Develop enumerations for the field: Experimental Strategy #178
• Experimental platform Develop enumerations for the field: Experimental platform #171
• Access Limitations Develop enumerations for the field: Access Limitations #172
• Access Restrictions Develop enumerations for the field: Access Restrictions #173

---

Pierrette's comments:
For Biospecimen/Assay/DataFile, we have discussed EDAM, OBI, etc. Sample Type, File Type, etc.

Currently, we have a controlled vocabulary, but using a pre-existing ontology might allow us to have more relationships/hierarchies, e.g., between Data Categories & Types. Maybe we could annotate Experimental Platform and then have a search widget for users to search parent categories?

[lopierra]

Application ontology will also be very useful for the new Study/Datatset fields. Please see the yellow highlighted cells in this sheet.

[lopierra]

Related to #159

[twhetzel]

Matt and I have discussed two options for including enums from ontology terms, (1) an explicit enumeration of terms in enum definitions or (2) referencing subsets of terms defined in an external Application Ontology.

Currently the decision is for (1) to define a fixed set of terms in the schema in advance because the LinkML docs says "this works well if the vocabulary is known in advance and it is a relatively small number of terms" [1].

A small example for sample type is in this branch

The notes from our discussion are in the INCLUDE gDrive here.

Two open questions are:

• Does the model have a display name field to customize the values?
• Can values be represented as a hierarchy?

[1] https://linkml.io/linkml/faq/modeling.html#how-do-i-constrain-the-value-of-a-slot-using-an-ontology-or-vocabulary