Modeling 'regulates o occurs in' annotation extensions about gocamgen HOT 6 OPEN

@ukemi Hopefully this looks right:

From model for MGI:MGI:101761. From the line you have above:

MGI     MGI:101761      acts_upstream_of_or_within      GO:0008284      MGI:MGI:3829847|PMID:18957199   ECO:0000315     MGI:MGI:2386412         20190426        MGI     occurs_in(EMAPA:32847),regulates_o_acts_on_population_of(CL:0000047)    contributor=http://orcid.org/0000-0001-7476-6306

from gocamgen.

Note that, if possible, we will try to model other chained annotation extension relations similarly.

from gocamgen.

@vanaukenk @ukemi I have this translation coded with the above example MGI:MGI:1929608 loaded into my USC server.

The code digs through the primary term's logical definitions in the ontology to find the regulated target term (e.g. regulation of mitotic cycle (GO:0007346) -> regulates some mitotic cell cycle (GO:0000278)).

However, I did find an example annotation of MGI:MGI:87859 to B-1 B cell homeostasis (GO:0001922) for which I could not dig the regulated term out of the ontology. Looking in ontobee, this term doesn't appear to have any logical definitions. I'm thinking either these LD's need to be plugged into the ontology or we adjust the extension rules TSV to limit regulates_o_occurs_in translation to certain BP terms that would have these LD's. Thoughts on handling this?

from gocamgen.

Seems like LDs should be added whenever possible ?

In this case there is a LD but it's weird:

'homeostasis of number of cells'
and (acts_on_population_of some 'B-1 B cell')

I thought for homeostasis we used the relation 'regulates levels of'; is there a reason why this one is not the case ?

Thanks, Pascale

from gocamgen.

Here is another example with a different chain for model: MGI:MGI:101761
MGI MGI:101761 acts_upstream_of_or_within GO:0008284 MGI:MGI:3829847|PMID:18957199 ECO:0000315 MGI:MGI:2386412 20190426 MGI occurs_in(EMAPA:32847),regulates_o_acts_on_population_of(CL:0000047)

Would be translated to: Hmg2a enables [molecular function] causally_upstream_of_or_within ([positive regulation of cell population proliferation] occurs_in [myenteric nerve plexus]) psoitively_regulates ([cell population proliferation] acts_on_population_of [neuronal_stem_cell]).

from gocamgen.

The example you have above is an anomaly. The parent regulation term regulates a biological quality/attribute, not a process. Any chain created for these terms would be added to the attribute, so this doesn't make sense. Attributes don't occur. This boils down to a curation error that I have now corrected. The homeostatic process occurs in the perotoneal cavity, the population of B-B cells doesn't occur there. When you find ones like this that have a regulates_o_blah chain and the subject isn't defined as 'occurrent regulates some occurrent', can you send them to me? They are curation mistakes. Hopefully there aren't that many.

from gocamgen.