query about PATO usage for decreased and absent processes

[mah11]

FYPO has a great many processual phenotypes, including some in which the occurrence of a process is decreased, and others where a process that normally occurs isn't observed at all.

Decreased occurrence is straightforward, using PATO:0002052 ! decreased occurrence, e.g.

decreased shmoo formation equivalent to 'decreased occurrence' and (inheres_in some 'mating projection assembly')

For mutants in which a process doesn't (detectably), we have logically defined "abolished process" terms using PATO:0001558 ! lacking processual parts, e.g.

abolished shmoo formation equivalent to 'lacking processual parts' and (towards some 'mating projection assembly')

It's not explicit in the logical def, but the lack-of-process is assumed to inhere in the cell.

Various PomBase curators and users would like to see ontology structures in which "abolished process" is a descendant rather than a sibling of "decreased [occurrence of] process", a la

decreased shmoo formation -- [is_a] abolished shmoo formation

Their rationale is that if a normal process isn't occuring at all, it's occurring less than in wild type.

First, is our approach to "abolished process" terms sensible, especially the current logical definitions? Or should we change -- or even abandon -- it?

Second, is there a good way to accommodate a 'decreased x' is_a 'abolished x' arrangement within FYPO? To date, we haven't asserted these links, and they don't get inferred from the PATO structure ('decreased occurrence' and 'lackiing processual parts' are in fairly different branches of PATO). But we could put something in FYPO that would lead a reasoner to infer decreased is_a abolished for process phenotypes, e.g.:

decreased biological process -- [is_a] abolished biological process

Assuming we retain 'abolished process' terms, is there any reason not to do this for FYPO?

I'm also enquiring on the phenotype-ontologies-editors mailing list.

Thanks, Midori

[cmungall]

1. We've stopped using towards so much, and are treating absence like a normal pato term, so here it would be just abolished and inheres in some P
2. We have a new QoP decreased efficacy that we use for terms labeled decreased P when P is a process and the definition pertains to how well the process produces its output. Previously we over-used rate but this should only be applied to cyclic or repeating Ps. Does this sound right for decreased shmoo formation
3. Given the above, I think the request to put abolished under decreased efficiency is sound. It's the processual equivalent of the rate limiting step. This would mean you get the desired FYPO graph edge by inference.

Regarding 1, I can help with the fypo edit file. I think you'll like this change overall.

For more details see https://github.com/gkoutos/pato/pull/49

[cmungall]

Apologies, I have probably created more confusion here in making a rushed reply. A few intertwined points:

• I was imagining 'abolished' was its own PATO class, as a processual analog to absence or agenesis. In fact 'abolished' is an exact syn for PATO:0000297 'arrested' which I think is confusion
• 'lacking processual parts' is a relational quality, for continuants we have stopped using these and gone back to 'absent', and specific child classes. We need to fully think through the parallel strategy for processes
• somehow the 'occurrence' classes missed our attention. We have barely used these for mp/hp

[mah11]

Hi folks,

Has there been any progress on this, new developments, etc? I'm finally getting back to thinking about it myself (for FYPO) -- I'd like to upgrade the "this process ain't happening" logical defs but want to do it so it's consistent with current PATO & phenotype ontology practices.

thanks, m

[mah11]

pinged again via obo-phenotype list

[ValWood]

HI, what needs to happen to discuss/resolve this issue? https://github.com/pombase/fypo/issues/4193#issuecomment-1203011404

Is there a reason not to relate "decreased" to "abolished" in PATO? Otherwise we need to instantiate all of these links manually.

thanks.

[sbello]

There is a pull request that @nicolevasilevsky is working on for adding the new classes https://github.com/pato-ontology/pato/pull/332

[ValWood]

Hi, any news on a connection between decreased and abolished yet? (and if you think this makes sense)

[sbello]

I've put this on the agenda for the next phenotype editors call https://docs.google.com/document/d/1WrQanAMuccS-oaoAIb9yWQAd4Rvy3R3mU01v9wHbriM/edit?usp=sharing on 11/17/22

[pfey03]

We use abolished for all that is absent, a biological process, a Molecular Function. Abolished is not arrested. We have arrested phenotypes for developmental stages

[ValWood]

If this is in response to

In fact 'abolished' is an exact syn for PATO:0000297 'arrested' which I think is confusion

I agree that arrested and abolished are not really synonymous.

Arrested in cell biology means stopped/attentuated and usually refers to temporal events like development @pfey03 refers to, or cell cycle progression. Arrested may or may not be reversible, but it means a process or event begins but cannot go beyond a certain point. To use arrested usually you would want to see a process halted at a specific time point, or at aspecific stage of a process.

Abolished is used when a function or a process is prevented from occurring at all.

We have problems because decreased and abolished are a continuum. Abolished is decreased with complete penetrance.

For example, we expect abolished macroautophagy to be a descendent of decreased macroautophagy via reasoning but we don't pick up these relationships.

I guess if some process is arrested (if cytokinesis is arrested at ring constriction), this could be described as arrested cytokines. However we would use abolished ring constriction because we want to be more precise (ring constriction is a part_of cytokinesis). Both would be correct, but it wouldn't occur to me to use 'arrested' here.