search

obophenotype / ncbitaxon

Build for NCBITaxon
BSD 3-Clause "New" or "Revised" License
25 stars 7 forks source link

Propose changing the label on the NCBI top level term to 'NCBI root' #10

Open cooperl09 opened 7 years ago

cooperl09 commented 7 years ago

In the Planteome project, we have found an issue since the NCBI top level term is called "root" and it conflicts with our PO term root (PO:0009005) in Protege, and looks weird on the AmiGO browser, .

Suggest that we need to add a modification of the NCBI top level term, which in NCBI is called "root" to be 'NCBI root'. I have contacted them and asked them to change it at their end, but they won't as we are the only ones who it is an issue for, apparently. See original post here: https://github.com/Planteome/planteome-ncbi-taxonomy/issues/1

ddooley commented 4 years ago

Or even spell it out to "NCBI organism taxonomy root" ?

cmungall commented 3 years ago

if we change it, it should be to "organism"

(although some subclasses are not organisms)

matentzn commented 2 years ago

Bringing this up again. Very in favour of that!

ehartley commented 2 years ago

image FYI, there are also environmental samples under root.

cmungall commented 2 years ago

@ehartley, exactly, many subclasses are not organism!

I think there is one coherent path forward, which is to inject COB organism as a superclass of cellular organisms and virus

After this we can simply ignore NCBITaxon:1 and its awkward confusing name. We would simply have all imports start from subclasses of COB:organism. We could also have a separate subset release of NCBITaxon with only these classes

mellybelly commented 2 years ago

We had created the union class in CARO to address this issue: http://purl.obolibrary.org/obo/CARO_0001010 Please do not label the union "organism" but rather something more expressive, else it is confusing and biologically incorrect. We spent a LOT of time discussing this previously, can COB just adopt what was decided in CARO or otherwise subsume some of these things? Things have not really changed and I would think the upper-level CARO ontology should be a reference here. @dosumis

matentzn commented 2 years ago

There are two issues mixed up here:

  1. "root" is a bad label and does not adequately capture the intention of the term. It also violates all sorts of OBO conventions. If its intention is to group all NCBI taxon terms, no matter what they are, we can just name it "NCBI-Taxon Root" as @ddooley suggests. We don't want to appear this class as root in any module, be it SLME OR otherwise.
  2. Whatever that class is called, it captures a variety of concepts that should be deal with independently, like Virus, and cellular organism. Therefore, we should, as @mellybelly says, not rename root to organism for sure. Instead, we should probably align all its directly children with COB 1 by 1 and ignore that class.