geneontology / go-annotation

This repository hosts the tracker for issues pertaining to GO annotations.
BSD 3-Clause "New" or "Revised" License
34 stars 10 forks source link

IEA with ARBA00026393 (should diphosphatase terms be phosphatases?) #4228

Open ValWood opened 2 years ago

ValWood commented 2 years ago

I am not sure if this is an ontology or an annotation issue:

Screenshot 2022-06-23 at 17 55 17

https://www.pombase.org/gene/SPBC17A3.03c

IEA with ARBA00026393

In GO di-phosphatases are not classed as phosphatases (presumably following CHEBI)

ValWood commented 2 years ago

@pgaudet

ValWood commented 2 years ago

I will transfer this issues, it seem like a GO issue?

pgaudet commented 2 years ago

This is intentional - phosphatases break monoester bonds https://enzyme.expasy.org/EC/3.1.3.- and diphosphatases break pyrophosphate bonds (P-P), https://enzyme.expasy.org/EC/3.6.1.-

ValWood commented 2 years ago

OK thanks, I will move this back to the annotation tracker.

PedroRaposo commented 1 year ago

Hello Val, I am not sure I understand this issue. The protein Q9UUF3 in PomBase is named as a phosphatase, and in UniProt we named it similarly. Also, UniProt gives the GO term GO:0016791 (phosphatase activity) because it is found in PTN001258896 (a node of the PANTHER tree PTHR31126). There are 28 Swiss-Prot with EC 3.1.3.- (with a phosphatase activity) (https://www.uniprot.org/uniprotkb?query=PTHR31126&facets=reviewed%3Atrue), except S. cerevisiae's P53965 (with 3.6.1.52). Most these EC 3.1.3.- annotation come from ProSite rule PRU10044 (https://prosite.expasy.org/rule/PRU10044) based on PS00383 signature. There are 2 options:

  1. Ask ProSite to review this signature, which might be giving the wrong EC to the Swiss-Prot entries. With that change, I can report this to PANTHER, to change the annotation of PTN001258896, and out protein Q9UUF3 would be automatically updated; or
  2. Ask curator to add the Rhea reactions mapped with the 3 molecular GO terms you have in PomBase:
    1. Add RHEA:22384 according to GO:0052845 (inositol-5-diphosphate-1,2,3,4,6-pentakisphosphate diphosphatase activity)
    2. Add RHEA:59500 according to GO:0106211 (inositol-5-diphosphate-1,3,4,6-tetrakisphosphate diphosphatase activity)
    3. Ask RHEA to create new reaction for GO:0052847 (inositol-1,5-bisdiphosphate-2,3,4,6-tetrakisphosphate 5-diphosphatase activity) as it there is no reaction associated with this GO. Then ask a curator to add this to the protein.

Let me know which option do you think is more reasonable

ValWood commented 1 year ago

Hi, The problem here is either i) a relationship missing from GO or ii) that although these inositol diphophatases are referred to colloquially as "phosphatases" they don't fit the phosphatase definition:

@pgaudet should a pyrophosphatase be is_a phosphatase?

Screenshot 2023-10-10 at 14 32 42
ValWood commented 1 year ago

I'm pretty sure the answer is ii)

For fission yeast I should annotate this product line more precisely as inositol pyrophosphatase

see https://pubmed.ncbi.nlm.nih.gov/37772819/ it is described as a x a cysteinyl-phosphatase-type inositol pyrophosphatas

phosphatase "Catalysis of the hydrolysis of phosphoric monoesters"

pyrophosphatase Catalysis of the hydrolysis of a pyrophosphate bond (diphosphate bond) between two phosphate groups.

PedroRaposo commented 5 months ago

Hi Val, I see this issue being unresolved. If the gene ontology itself needs to change, maybe we can assign this to someone from GO. Who should be the best person to assign this to?

ValWood commented 5 months ago

Based on the above discussion I think the problem is the mapping/inference not the ontology

PedroRaposo commented 5 months ago

I can see the GO term GO:0016791 (phosphatase activity) does no longer come from ARBA ARBA00026393, instead if comes from a mix of other IBA (Inferred from Biological aspect of Ancestor) evidences. If anything, what needs to be updated are those evidences. Do you agree?

image

ValWood commented 5 months ago

agreed changing to PAINT

@marcfeuermann

marcfeuermann commented 5 months ago

PTHR31126 is a big family containing several kinds of phosphatases, that's why they all have been annotated to te root with the term "phosphatase activity". It allows to give a MF to the plant members for which we do not know yet the exact subtrates. As an example, "phosphatidylinositol-4-phosphate phosphatase activity" has parent "phosphatase activity" but "inositol diphosphate tetrakisphosphate diphosphatase activity" is not a "phosphatase activity". A diphosphatase should be a phospatase, no ? This clearly looks like a hierarchy inconsistency, not a PAINT problem, All those are in my opinion "phosphatase activities". I could remove the annotation but this is hiding the dust under the carpet. The solution is to clean the ontology. This is it for my input. Thanks a lot, Best regards, Marc.

ValWood commented 5 months ago

Actually, I agree, it does seem weird that diphosphatases are not classed as phosphatases @pgaudet

pgaudet commented 5 months ago

Hi @ValWood see my comment above: https://github.com/geneontology/go-annotation/issues/4228#issuecomment-1167208310

This is not the same reaction: phosphatases break monoester bonds https://enzyme.expasy.org/EC/3.1.3.- and diphosphatases break pyrophosphate bonds (P-P), https://enzyme.expasy.org/EC/3.6.1.-

so they cannot be children of each other.

marcfeuermann commented 5 months ago

Hi all, I understand the distinction, but a personal thought that commits only me, it shows how far we've gone in the complexity of the ontology. These enzymes are part of the same family, their purpose is to remove a phosphate from closely related compounds, but we can't annotate them in the same way, it's so weird. Most curators and, I think, users (we often forget that we're making ontology for them), just don't get it. One solution I could see would be to distinguish between monophosphatase and diphosphatase, both of which are children of phosphatase . Best regards, Marc.

ValWood commented 5 months ago

Marc's solution sounds better, I think these phosphatases should have a common parent...

marcfeuermann commented 5 months ago

To add a little water to my mill: what about the kinases that catalyze reverse reactions ? No one sees any problem with them, yet the problem must be essentially the same, right ?

Antonialock commented 5 months ago

@marcfeuermann what's an example of a kinase that catalyses a reverse reaction? I donät think I'm familiar with these and would like to look one up!

marcfeuermann commented 5 months ago

@Antonialock... my previous message was probably not very clear... All kinases catalyze reactions that are reverse to phosphatases. So, as different types of phosphatases break monoester bonds or pyrophosphate bonds, different types of kinases create these monoester bonds and pyrophosphate bonds. And all these activities are considered as kinase activities.

Antonialock commented 5 months ago

Ah that makes more sense! Thanks