biolink-model icon indicating copy to clipboard operation
biolink-model copied to clipboard

Exposure outcome modeling

Open karafecho opened this issue 5 years ago • 5 comments

A clear and concise description of what the request is.

I want to represent exposure outcomes in Biolink.

What working group (or team) did this request originate from?

The need for representing this type of data originated from the Translator 2020 exposure-outcome breakout session on 9/23/2020..

Note:

This is relevant for members of NCATS Translator.

Describe the solution you'd like

We'd like to define new outcome semantic types (other than phenotype or disease) and identify a general approach for representing outcome metadata.

Additional information to support this request

image

**Tag relevant members for discussion

@karafecho @diatomsRcool @stevencox @xuhao

karafecho avatar Sep 24 '20 18:09 karafecho

Would biolink:DiseaseOrPhenotypicFeature work for representing outcomes? Or is that too misleading in terms of representation?

Another way would be to have biolink:Outcome as a sibling to biolink:DiseaseOrPhenotypicFeature

deepakunni3 avatar Sep 28 '20 17:09 deepakunni3

The challenge is that outcomes such as 'ED visits' or 'hospital readmission' or 'surgical intervention' do not really map to DiseaseOrPhenotypicFeature. 'Sibling' might work, but I'm wondering if outcomes are better modeled as 'outcome events', akin to 'exposure events'? I'm by no means an expert, however, so my suggestions may not make sense.

Thanks for looking into this!

karafecho avatar Sep 28 '20 21:09 karafecho

During Friday's meeting of the Clinical Data WG, I put forward another common outcome metric: death. Anne T. clarified that this is currently treated as a 'PhenotypicFeature' in Biolink. The group discussed whether that makes sense or not, as 'dead' ('alive') seems more like a 'PhenotypicFeature', whereas 'death' seems more like an outcome or event. We also discussed whether outcomes could be modeled as a variant of 'exposure events'. No decisions were made, as we are planning to continue this discussion and explore other options in Biolink.

Just passing this info along to keep you informed and see if you have any thoughts to add...

karafecho avatar Oct 05 '20 16:10 karafecho

was this resolved?

nlharris avatar Jun 08 '21 16:06 nlharris

This is being actively worked on by Matt, Richard, myself, and others on behalf of the clinical data committee. Requirements gathering is complete; implementation has yet to be initiated.

karafecho avatar Jun 08 '21 18:06 karafecho