1a6: Ontology and model identification and selection for particular data / entity types

[jmcmurry]

This can be closed when a formal recommendation has been made about which ontology should be used for each data element in the harmonized model here.

[ShahimEssaid]

This primarily sounds like a modeling WS task rather than terminology WS, and should go on that board or at least be on both. Also, this work is in progress and we don't need to discuss in detail on the next call. We have a plan and will be busy for few weeks. We can revisit this in 2-3 weeks.

[nicolevasilevsky]

Which is the modeling WS, @ShahimEssaid? Please feel free assign/reassign labels here too. :)

[gaurav]

@nicolevasilevsky I guess that would be https://github.com/orgs/cancerDHC/projects/1?

[bfurner]

This issue depends on having a fairly complete harmonized model, at least at the entity and property levels which is not the current state. Presently, the ADM represents a minimally refactored covering model of the source node models and we are currently working on:

1. mapping ADM entities and properties to BRIDG, and
2. evaluating the utility of representing ADM entities and properties within the FHIR modeling language

We are not yet at a point where we are harmonizing at the value set or ontology level

[jmcmurry]

We may want to consider a bottom-up approach where datatypes are given a provisional recommendation for ontology and we revisit for the model as a whole? I'm assigning this ticket to Brian and Matt for now.

[monicacecilia]

@bfurner please indicate whether releasing v1 of CRDC-H means that this task can be considered complete, or whether it should continue to move forward (and if so, please indicate through which phase (i.e., through Phase III or through Phase IV)). Thank you! 🌷

[bfurner]

@monicacecilia this will carry forward until, at a minimum, the CRDC-H covers all of the major subdomains in question, so v1 isn't completion of this. i hope that by end of Phase III we would have the modeling provisionally completed for this. the question of binding the model to specific terminologies, which is what this ticket is really about AFAIK, is a cross-cutting concern with the model and terminologies team so this should be coordinated. modeling has been relying upon the terminologies team to do the value set harmonization and mapping to terminologies. ultimately, this work will have to continue for as long as the source models continue to evolve and so really will likely extend through Phase IV

[monicacecilia]

Thank you for addressing this, @bfurner. This clarifies things for me. 🌷