Proposal: reuse upper level OBI classes

[cmungall]

OBI has classes for Assay (test), biopsy, etc. These could be used in dosdp files and used as upper level classes in MAXO

[LCCarmody]

Looking at OBI, there are some relevant terms under 'assay', including imaging assay, hematology assay, and imaging assay (to name a few).

Biopsy isn't in the ontology as a procedure, but only the resultant, "pre-mortem specimen". This might be helpful in describing the biopsy but is not the biopsy itself. I suppose we could request to add these terms?

There isn't much hierarchy there, either. Would it be better xref?

[cmungall]

I think it would be good to reuse classes. You can request new upper classes from them. Be good to align terminology. If you think the term biopsy indicates a material and not a process then you can bring this up on the obi tracker

…

On Mon, Apr 29, 2019, 10:39 Leigh Carmody ***@***.***> wrote: Looking at OBI, there are some relevant terms under 'assay', including imaging assay, hematology assay, and imaging assay (to name a few). Biopsy isn't in the ontology as a procedure, but only the resultant, "pre-mortem specimen". This might be helpful in describing the biopsy but is not the biopsy itself. I suppose we could request to add these terms? There isn't much hierarchy there, either. Would it be better xref? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#30 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAAMMOL34EWE6MJZKVMUR2DPS4XDHANCNFSM4HJETYDQ> .

[matentzn]

Usually I am 100% with you @cmungall on re-using stuff; in this case I would be more inclined however to create a separate alignment. @LCCarmody has done some good work to convince me that the hierarchies of the ontologies she wants to pull together, including NCIt and OBI, do not reflect the clean groupings we are looking for now with MAXO. They are not UBERONs either where we can just go and fix stuff when we need to. I think Leigh will, in her engineering practice, carefully document alignments with other ontologies. But the Medical actions space is a similar chaos to the disease space, and, at least for the first 3 levels of the hierarchy, I would prefer we were independent and drive integration with bridges ala UBERON.

[cmungall]

This seems sensible, it would just be good to articulate some of the reasons not to reuse - and feed it back. In theory OBI and other OBOs should be open to change. It would be very useful for me to know areas in which these OBOs don't work. As it happens I have been looking at OBI "specimen" class lately and I had some questions/concerns about it.

By all means let's start with some equivalence axioms between the named classes (or xrefs if you must..)

[bpeters42]

This ticket makes it pretty clear that MAxO has no intention of conforming with OBO foundry principles; why are you applying for a purl then?

[ddooley]

On our call today we discussed "biopsy" that Chris raised in (obi-ontology/obi#1072), and see that it would be a specimen collection process. It currently is listed under ERO, so I am following up with ERO to see about importing it and its subclasses into OBI.

I'll just add that it is incumbent on us at OBI to understand any class structure issues that make MAXO curators shy of integrating ... we would appreciate knowing what barriers to adoption you see in particular.

[LCCarmody]

Hey @ddooley we are happy to align on biopsy. Please see: obi-ontology/obi#1072 (comment)

[LCCarmody]

Add Collecting with a swab: use for swab collection pattern http://purl.obolibrary.org/obo/OBI_0002600