Ontology meeting 2021-09-27
- Group members: Pascale, Karen, Harold, Raymond, Peter, Jim, Tanya, Kimberly, David, Chris, Paul
- Present:
- Regrets:
Announcements
Agenda
GOC meeting - Oct 12-14
Agenda: https://docs.google.com/document/d/15_TIWfnu2X_3CVzdcou14PyV4Ff8F9Hz8WSASi1bsaE/edit#
Suggestion to have a group discussion (breakout?) with people developing external ontologies to discuss expectations, semantic equivalence, etc
New Taxon constraints
https://github.com/geneontology/go-ontology/issues/16298 Any new item to review?
methotrexate terms
Obsoletion proposal for methotrexate terms Quick summary from Karen
The issue to discuss is the fact that the term “response to methotrexate” is imported into the Experimental Factor Ontology (EFO) and used by curators at the Genome-Wide Association Studies (GWAS) catalog, and how to proceed, whether to continue doing what was agreed upon 5 years ago:
place them in the do_not_manually_annotate subset, and add a definition comment "Note that this term is in the subset of terms that should not be used for direct manual annotation of gene products. It was created to be used for cross-referencing by other ontologies. Direct annotations to this term may be amended during annotation QC."
OR If GO still wish to go ahead and obsolete GO:0031427 'response to methotrexate', the alternative option for EFO would be to mint a new EFO term with the same meaning, use it to replace the GO term, and alert GWAS curators
=> DECISION: Notify annotators and people using GO (external ontologies) way in advance (6-12 months) of the upcoming obsoletions, and proceed, at least for 'response to drug' and other non-endogenous chemicals
Multiorganism processes
Two issues
1. 'symbiont regulation of host processes' terms
right now some (but not all) link to the 'normal' process, for example (made up), 'inhibition of host transcription' could have in its subclasses or LD, 'negatively regulates some transcription, DNA templated'
2.
Current pattern is
- 'modulation of process of other organism'
- 'modulation of process of other organism involved in symbiotic interaction'
- 'modulation by symbiont of host process'
- 'modulation by host of symbiont process'
- 'modulation of process of other organism involved in symbiotic interaction'
+ suppression/induction children
Issues are : - structure is complex - forces to create 2 independent branches for (intra-organism) symbionts and others (snakes, frogs, etc) - not always very relevant, usually by knowing the pair of organisms you know the relationship - there is probably a better way to capture this information, using context (extensions) that would represent inter-organism relation(host/symbiont, prey/predator, etc) - proposal would be to collapse all terms to “modulation/suppression/induction of x in another organism”
MF regulator activity
As we (try to) remove BPs of type 'regulation of MF', people are asking for 'x regulator activity', but in many cases it is not clear whether it is true regulation, or just that the protein is 'required for' the activity (whatever that might mean) see NTR: tRNA ligase activator activity, citing PMID:30943413
Protein-containing complex localization & protein localization
Inconsistent 'design patterns'
Amino acid conjugates
https://github.com/geneontology/go-ontology/issues/22096
Ontology editors documentation
- Review new term creation documentation: Guidelines_for_new_Molecular_Functions and Principles_for_creating_a_GO_term
- Review cross-reference documentation: Guidelines_for_'database_cross_references'_and_'definition_database_cross_references'
- Review protein binding term documentation http://wiki.geneontology.org/index.php/Guidelines_for_new_Binding_terms
- Optional: Review Chris's slides on aligning editor/contributor docs across all of OBO: https://docs.google.com/presentation/d/1YyWYSetDML3nhNgMUmcrsw80Xx1Uf1n1dWI-A20PMa8/edit#slide=id.p
- NAD/NADP: do we keep a grouping term even if all annotations can be classified to NAD or NADP-dependent? for eg GO:0003845 11-beta-hydroxysteroid dehydrogenase [NAD(P) activity] - all known enzymes have known specificity, see review ticket
- added 'do not annotate'
- DECISION (last call): This is what we will do
- Need to add to documentation