Ontology meeting 2024-02-12

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

Note that Peter will join at 9h30

Discussion points

Obsoletion announcements

  • one of the items we check is whether is GO term is in a subset; note that we dont need to announce that terms are present in the 'antislims' (do not annotate)

Automatic mappings

From Chris in this ticket: https://github.com/geneontology/go-ontology/issues/26941

 skos:exactMatch is transitive so we can assert inferences from this. GO:x
  exactMatch RHEA:y exactMatch MetaCyc:z ==> GO:x exactMatch MetaCyc:z
  • Can we have a script that handles this?
  • RHEA Exact matches?
    • Anne's reply: Unfortunately the mappings between RHEA and METACYC, EC, KEGG are NOT exact matches for few cases. A search in the Rhea database (internal):
      • 324 MetaCyc reactions are linked to more than one Rhea
      • 116 KEGG reactions are linked to more than one Rhea
      • 38 Rhea are linked to more than one KEGG
      • 87 Rhea are linked to more than one MetaCyc
    • Strategy: look for recirprocal 1:1 mappings; apply automatically; others: apply BROAD NARROW match automatically?


Currency chemicals

https://wiki.geneontology.org/Currency_chemicals Obtained from Reactome, see https://github.com/geneontology/go-ontology/issues/26900

  • The list currently contains chemicals found in RHEA (pH7.3 list)
  • Questions from Peter:
    • Should we have OH-, with consider H2O
    • Should we include the non-pH7.3 version ?
  • Questions from Kimberly:
    • We need to clarify exactly what input/output relations will be on the Noctua annotation UIs and why
      • New standard annotation UI: has primary input/output for MF and BP
      • VPE: also primary input/output on MF and BP?
        • For imported models that might be used as templates for manual models, do we want the option to edit to indicate primary vs not-primary?
        • If MF terms have 'has participant' lists, do we want those available for editing to primary vs not-primary?
        • It would be good if curators don't have to copy paste ChEBI ids from another source, or potentially choose the wrong form of a chemical when we already have that information in the ontology.

User request: gasoreceptor activity

https://github.com/geneontology/go-ontology/issues/26963#issuecomment-1930726894

Everyone agree that this is out of scope; we don't to capture which chemicals are in gaseous form at room temperatures.

Blocking issues

ongoing tickets

Review design patterns (continued from previous calls)

Repo: https://github.com/geneontology/go-ontology/tree/master/docs/patterns

transport

signaling receptor activity by input

ligand incorrectly identified as the input (GH ticket)

      • Broader question: do we want all ligands as MF? Or should these move to x signal transduction, and keep a single activity, 'signaling receptor activity'?
        • Noting that most 'signal transduction' (BP) children have no logical definition. How about, for example:
 GO:0038027 apolipoprotein A-I-mediated signaling pathway
 'cell surface receptor signaling pathway'
  and ('starts with' some 'signaling receptor activity')
  and ('has small molecule activator' some 'apolipoprotein A-I')
  • Noting also that there is no 'small molecule' in ChEBI; so any PRO is a ChEBI:chemical entity
  • Discussion and AI
    • We do want to keep the receptor activities and use these to logically defined the signaling pathways
    • Remove relationship: signal transduction part of cell communication; this is not always true, since there are intracellular signaling pathways
    • Add new DP for signal transduction

involved_in_x_y

Other tickets/projects