Ontology meeting 2022-11-28

From GO Wiki
Jump to navigation Jump to search
  • Group members: Pascale, Harold, Raymond, Val, Edith, Peter, Jim, Tanya, Karen, Kimberly, David, Chris, Paul
  • Present: Pascale, Harold, Raymond, Val, Edith, Peter, Jim, Tanya, Karen, Kimberly, Chris, Paul
  • Regrets: David

Announcements

This is the second-to-last call of the year. Next week (Dec 5th) we will talk about cell differentiation & children term definitions, Dec 12th is the no GO calls week, and Dec 19th Pascale (and others?) is on vacation.

Improving workflow for obsoletions

Obsoletion Ticket templates

Are term merges allowed any more in any circumstances or is it *always* obsoletion?

  • If terms are not used by external ontologies, can they be merged?
  • Issue is this is more work for ontology editors + not sure all announcements are required// are we not spamming people?


GO update paper 2023

https://docs.google.com/document/d/1jaGHywr2_DrflIDUqs0iebmfSRssTcOpcNtIsfFg-Dw/edit# See section on ontology; we explained many reasons of changes, so users better know what to expect

Projects

unnecessary grouping terms in the chemical homeostasis branch

https://github.com/geneontology/go-ontology/issues/24455

and more generally, the correct biological level of classification for chemicals:

  • ion (ie , monoatomic ions)
  • amino acid
  • carbohydrate
  • lipid
  • nucleic acid
  • other molecules that dont fall in these classes:
    • water
    • ammonium
    • phosphate
    • sulfate
    • ascorbate
    • possibly others

So, Paul and Pascale proposed obsoletion of terms such as

  • (cellular) mon(di/tri)ovalent inorganic ion/anion/cation homeostasis

with replaced_by ion/anion/cation homeostasis, implying 'monoatomic ion'

  • Changed the logical definitions of
    • 'ion homeostasis'
    • 'anion homeostasis'
    • 'cation homeostasis'

from using chebi 'ion', 'anion', 'cation', to 'monoatomic ion', 'monoatomic anion', 'monoatomic cation', respectively

>> probably need to update the definitions (and term labels??) to reflect this. This needs to (1) be clear and (2) be easily understandable by users

Need to do the same in other branches where 'ion' is a grouping class':

  • 'GO_0043167 ion binding'
  • 'GO_0005216 ion channel activity'
  • 'GO_0034702 ion channel complex'
  • 'GO_0006811 ion transport'

(in new tickets)

Obsolete BPs "regulation MF"

https://github.com/geneontology/go-ontology/projects/34?add_cards_query=is%3Aopen OK to go ahead with obsoletion?


Suggestions for 'homeostasis at the organism level in multicellular organisms'

  • 'Circulating levels' used in animals; does that make sense in pants? Tanya -> no
  • suggestions (by email):
    • intra organismal x homeostasis
    • intra multicellular organismal x homeostasis
    • organismal x homeostasis (+ restriction of do not annotate single celled organisms)


When it is appropriate to add vague strings as synonyms

Val

  • With the Elixir biocuration Focus group we are trying to establish what types literature mining tools and outputs are useful for Biocurators (I think the main purpose is mainly to influence developers of such tools under the Elixir umbrella to support biocurators needs, but obviously these will hopefully be useful general guidelines).
  • I was looking at some GO text-mining this morning and complaining that the Europe PMC contributing text-mining tools were using substrings: http://europepmc.org/article/MED/33579781 (see the GO "annotations")
  • I made this statement
  • Subtrings of GO terms should not be identified. “Uptake” and “synthesis” “biosynthesis” do not exist as GO terms independently. However when I checked "uptake" is a broad synonym of GO:0098739 import across plasma membrane. Should we be adding synonyms like this (with any scope?)
This has nothing to do with substrings AFAICT. Like many text mining groups, this group has not engaged with the ontology community and are using synonyms naively, in this case, weighting non-exact synonyms the same as exact synonyms. We should write documentation for GO like this https://github.com/obophenotype/uberon/wiki/Using-uberon-for-text-mining and also get this into OBO-central documentation. At the same time we should pay more attention to our own synonyms. We have "synthesis" as EXACT for biosynthetic process which is dubious. We also have "apoptosis" as a NARROW synonym which is wrong. It should be very rare to have standard molecular biology terms be anything other than primary labels or exact synonyms. Cjm (talk) 11:21, 28 November 2022 (EST)

Following up on prevous items

Discussions arising from BP refactoring

Cellular component Assembly

  • AI: query genes annotated to both assembly of a complex and part of the same complex.
  • AI: improve definition to clarify that the annotated genes must mediate the process, not just be a structural constituent

NAD, NADH, NADPH processes logical definitions inconsistencies

Jim https://github.com/geneontology/go-ontology/issues/24279

regulation of pH

Jim https://github.com/geneontology/go-ontology/pull/24256/files

cell morphogenesis

Causing DP issues: https://github.com/geneontology/go-ontology/issues/24299

Could we say that for cells, morphogenesis is the same as differentiation?

regulation of biological quality

Add axioms for regulation of biological quality

Retrieved from "https://wiki.geneontology.org/index.php?title=Ontology_meeting_2022-11-28&oldid=83764"