Settle property / class name: Cohort vs Study Group in core-im / caf profile

[mbrush]

In a StudyResult, the core IM uses the term 'Study Group' rather than 'Cohort' or 'Population':

StudyResult.studyGroup property definition: "A structured description of specific population of subjects interrogated in the Research Study to produce the subset of data captured in the StudyResult.".

StudyGroup class definition: "A collection of individuals or specimens from the same taxonomic class, selected for analysis in a scientific study based on their exhibiting one or more common characteristics (e.g. species, ethnicity, race, country of origin, clinical history, age, gender, geographic location, income, etc.) May be referred to as a 'cohort' or 'population' in specific research settings."

The term "study group" was chosen here as a more generalizable name - to account for study groups that are not cohorts in the strict human-centric / clinical sense.

In the CAF profle , the term 'cohort' is used for the property, rather than 'studyGroup'. Also, there is no stand-alone 'Cohort' or 'Population' or 'Study Group' class defined/used in the Profile (rather, such a class is implied by the properties defined under the object taken by the 'cohort' property).

This ticket specifically concerns the most appropriate name for this concept in the va-spec models (StudyGroup vs Cohort vs Population vs . . . )

Issue #142 discusses the lack of explicit use of / reference to a Cohort/StudyGroup class in the caf profile.

[larrybabb]

@mbrush I think the right solution here is to rename the CAF.studyGroup attribute to CAF.cohortStudyGroup in the standard profile and if we wish to shorten that to CAF.cohort in the implementation profile for gnomad's CAF profile then we will do that.

Agreed?

[larrybabb]

@mbrush this is part of issue #141 and I have applied this to the latest refactoring of the core and profile schemas. Please review and close this if it meets your expectations.

[mbrush]

I see what you did here, extending the core-im studyGroup property into cohortStudyGroup in the CAF profile. I am fine with this - but may also be ok with just renaming the property cohort in the profile if you want.

I suspect you named it cohortStudyGroup to be consistent with the pattern of naming we used when extending SPOQ properties - where we want to keep the original name of the core-im property in the extended property name. I'd like to apply this convention toward all property extensions where it makes sense to do so. But as I've said, I am less concerned with this convention when extending non SPOQ properties.

In this particular case cohortStudyGroup seems like it might introduce confusion - like we are talking about a study group that is part of a cohort or something. And the idea that a 'cohort' is another name for a study group in specific settings is already introduced in the definition of the core-im studyGroup property. So here, simply calling this cohort as was in the original caf definition may be fine by me.

[ahwagner]

Agree with @mbrush; this is more confusing with StudyGroup than without, and recommend revert to cohort.

@larrybabb