Design pattern for vertebral elements: numbered and region based #82

Open
cmungall opened this Issue Sep 18, 2012 · 2 comments

Projects

None yet

1 participant

@cmungall
Member

prior to merge, TAO had the following structure:

  • CARO:0000003 ! anatomical structure
    • CARO:0000024 ! compound organ
      • VSAO:0000128 ! skeletal element
      • VSAO:0000184 ! vertebral element
      • TAO:0001189 ! vertebra
      • TAO:0001980 ! vertebra 6
      • TAO:nnnnn ! vertebra N

Note the def:

  • TAO:0001189 ! vertebra ``Endochondral bone that forms around the notochord and is part of the vertebral column.`

i.e. the hierarchy doesn't place this under bone but the def does. The uberon merge was based on the text def, so we now have:

(note polyhierarchy)

What was the original intent?

Either way, to be maximally useful (e.g. shark annotation), perhaps "vertebra N" should be placed under UBERON:0010913 "vertebral element", and renamed "vertebral element N" in ext?

Alternatively we could materialize the cross product of {element, bone, cartilage} x {1,2,...,N} but this seems over the top.

The disadvantage of making the classification more generic is that the vertebrae will not show up in a naive query for bones. However, if the query engine uses an OWL reasoner, it can be made to return valid and complete answers where we have taxon GCIs of the form "every vertebra in taxon T is a bone"

Comments?

Note even though this is in ext, it may make sense to make the changes en masse.

Note also that we have a separate parallel mammal-centric numbering in core based on C,T,L,S etc. To be discussed in Asilomar...

@cmungall
Member

Comments from David B:

  • "in salamanders there is an argument that vertebral development is perichondral and not preformed in cartilage. In this case, would it be preferable to have it as a "skeletal element" instead of specifically an "endochondral element"?"
    From Wasila:
  • "it seems more appropriate to go with 'skeletal element'. We cite other examples of variable development of the vertebrae in the vsao manuscript."
@cmungall
Member
  • generalize existing "vertebra N" classes in ext to be vertebral elements (which is a subclass of skeletal element)
  • use unambiguous primary labels: V element, V bone, V cartilage, etc
  • the existing replaced_by mappings from TAO to Uberon can remain in place
  • add taxon GCIs to indicate which vertebral elememts are endochondral
@cmungall cmungall changed the title from vertebra N [ext] to Design pattern for vertebral elements: numbered and region based Nov 27, 2015
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment