Annotating to a 'novel' concept

[raymond91125]

moving from https://github.com/geneontology/noctua/issues/365

I am trying to use Noctua to annotate cell/anatomy function. My first test case is a statement: "The role of the excretory cell is probably osmotic/ionic regulation and waste elimination, analogous to the renal system of higher animals." Just for the first part, I think I need a concept like 'regulation of organism osmolarity'. Is there a way that I can construct this concept within Noctua? I have not found a similar concept in GO.

@cmungall [your mf]--regulates-->[osmolarity]--inheres_in-->[multicellular organism] you should be able to find similar concepts in GO (use go-plus or the editors file and look at it in Protege). The RoBQ terms use the same pattern (using pre-composed terms from OBA, but these have OWL equivalence axioms using this pattern)

@ukemi

Hi Raymond,

I'm pretty sure we can do these kinds of things. Maybe the route of relating the cell and the process would be the way to go.

worm excretory cell capable_of some multicellular organismal water homeostasis.

or stronger worm excretory cell enables some multicellular organismal water homeostasis

Would the homeostasis GO terms work?

-D

[cmungall]

I like @ukemi's idea - cut straight to a homestasis class.

Of course, if there is a named cell type here and that cell type evolved to regulate water homeostasis then the best place for this axiom is WBbt. You'll need to set up imports

[raymond91125]

Kimberly suggested a similar approach as David's. I actually prefer the annotation resembles as closely as possible the author's statement. I did manage to enter a model http://noctua.berkeleybop.org/editor/graph/gomodel:57ec3a7e00000027. Yes, point taken about adding axiom to WBbt.

[cmungall]

the inheres-in between organism and excretion is not correct (you'll soon be forbidden from doing this)

cel --[enables]-->excretion--[regulates]-->os.ion.homeostatsis--[occurs in]-->organism

you don't need inheres-in if it's embedded with in the GO process, as it is here

[raymond91125]

Actually, is there a way to use Noctua to add class axioms to an ontology, e.g. WBbt, as opposed to annotate individuals?

[cmungall]

Not yet. What is your use case?

There are external tools you can use to generalize an instance graph to a class graph. For example, you may like to do a subset of the wbbt lineage in noctua where you have the graphical editing, export the OWL, and run owltools --abox-to-tbox. At the moment you'd still have to go in and add labels.

We can also imagine wanting a pathway described in noctua to have its own class in the ontology, but the hooks for doing this are not there yet.

[raymond91125]

The use case is this one above where I used Noctua to annotate known functions of a worm cell type. As you suggested that I should add the axioms to WBbt. But it doesn't seem straight forward to convert Noctua-generated owl to something I can just slurp into a wbbt.owl.