Can anatomical entity patterns be used for fungal anatomy?

[jseager7]

While trying to map terms from the Pathogen-Host Interaction Phenotype Ontology (PHIPO) to uPheno patterns, we've run into a problem with phenotypes for fungal species.

As an example, we decided that the PHIPO term 'reduced hyphal branching' was a good enough fit for the pattern 'abnormally decreased number of anatomical entity'. We were planning to use 'hypha' from the Fungal Anatomy Ontology as the variable (FAO:0001001), until we realised that the pattern only accepts terms from Uberon, and Uberon won't accept terms from FAO:

Uberon editors will not take requests for terms from outside of the Metazoa. Plant anatomy term requests should be directed to the PO, fungal anatomy terms to FAO.

(Source: https://github.com/obophenotype/uberon)

So it seems like the pattern 'abnormally decreased number of anatomical entity' and other related anatomical entity terms really mean 'abnormally decreased number of metazoan anatomical entity', which seems unnecessarily restrictive.

Is the only solution to write our own equivalence axioms (or patterns) that permit terms from FAO instead of Uberon?

[matentzn]

@jseager7

This is a great question. Since Fission Yeast is already in uPheno, it is fair to say that while "uberon" is metazoan, uPheno is probably better conceptualised as "scope-less".

So I would suggest we expand the scope to FAO and PO as well.

The integration will be a bit more tricky, but maybe there does not have to be integration for now on macroscopic (super cellular) structures.

The way Uberon anatomical is integrated into COB, it is fair to say that this term, while Uberon itself is metazoan, the _class) anatomical entity (and its immediate subclasses) apply to all super cellular structures (including single cell organisms and multi-cellular structures).

[rays22]

While the concept of an 'anatomical entity' technically encompasses diverse taxa including Fungi, Metazoa and Viridiplantae, we must exercise considerable caution when attempting to group phenotypes from fundamentally non-homologous structures.
The developmental processes governing anatomical structure formation in these taxonomic groups emerged independently in evolution. Their underlying genetic programmes diverged so substantially that any putative commonalities hold negligible practical value for interpreting most phenotypes across these phylogenetically distant taxa.

In other words, grouping together fungal, plant and animal anatomy phenotypes would make it more confusing than helpful to interpret them across these distant taxa.

Before consolidating fungal or plant anatomical entity phenotypes with those of animals using the 'abnormal anatomical entity' uPheno patterns, I would strongly suggest establishing distinct ontological branches specifically for fungal and plant anatomical entities. This approach would ensure that semantic similarities more accurately reflect true biological relationships, rather than imposing artificial taxonomic groupings that obscure fundamental biological differences. An example would be how 'fungal cell' is made explicitly distinct from other eukaryotic cells.

[matentzn]

Before consolidating fungal or plant anatomical entity phenotypes with those of animals using the 'abnormal anatomical entity' uPheno patterns, I would strongly suggest establishing distinct ontological branches specifically for fungal and plant anatomical entities.

100% - Uberon, FAO and PO are not integrated - they would exist in separate branches only. uPheno would simply ensure they use the same pattern, but they wont group anything together.

[rays22]

Before consolidating fungal or plant anatomical entity phenotypes with those of animals using the 'abnormal anatomical entity' uPheno patterns, I would strongly suggest establishing distinct ontological branches specifically for fungal and plant anatomical entities.

100% - Uberon, FAO and PO are not integrated - they would exist in separate branches only. uPheno would simply ensure they use the same pattern, but they wont group anything together.

If same patterns are used, but fungal, plant and animal anatomical entity phenotypes are not grouped together in uPheno too closely, because fungal and plant anatomical entities were different branches under the generic 'anatomical entity' class, then using terms from FAO instead of Uberon should address my concerns. I think the remaining issue is to integrate fungal structure with Uberon 'anatomical entity' somehow.

[matentzn]

I think the remaining issue is to integrate fungal structure with Uberon 'anatomical entity' somehow.

@jseager7 do you have any relationship to FAO devs? Could you make an issue to ask if they'd consider adding this as a top level alignment and tag me there?

[jseager7]

@jseager7 do you have any relationship to FAO devs? Could you make an issue to ask if they'd consider adding this as a top level alignment and tag me there?

Unfortunately not, but I might have connections to some people who do. I'd be happy to make the request but I'm not really sure what I should be asking for since I don't know much about the technical side of phenotype alignment. Unless you just want me to repeat what I was asking for here.

There's a possibly related issue already on the FAO tracker, though it's been inactive for years: obophenotype/fungal-anatomy-ontology#17

[ValWood]

PomBase said they would do this temporarily when FAO was left without an owner.

I haven't edited it yet but if you open a ticket.
I don't have the repository checked out, and I have no idea if there is an official release pipeline!

[matentzn]

I can help with the release if someone tags me