Ontology meeting 2015-10-08
Attendees: Melanie, Paola, DavidOS, DavidH, Harold, Chris, Tanya
Minutes: Tanya
Mobile protein/RNA
Additional discussion based on community feedback.
SO not the ideal solution: https://github.com/The-Sequence-Ontology/SO-Ontologies/issues/333
Ideas for term enrichment:
there are TE tools that don't use GO or don't use GO exclusively, GO is not the only way to enrich ToppGene - web based tool , not GO only http://www.broadinstitute.org/gsea/msigdb/index.jsp Ontologizer might work as a neutral site, + a separate GAF (?) create a 'gene set' tagged with a paper, see: some gene sets in the GSEA Broad collection that are grouped by paper
Ideas for annotation:
PATO:mobile + gp identifier
gp start_location anatomical part 1end_location anatomical part 2
For FT: 'cell fate commitment involved in positive regulation of flower development' IGI FD has input 'shoot apical cell' has output 'floral meristematic cell' located in nucleus part of 'shoot apical cell'
possibly wait till more is known about the genes in this dataset and then annotate function/process?
Also points out the need for (even) more user outreach and education, what is the scope of GO and what falls outside of that scope.
AI: get advice from PaulT, since he heard this discussion last week, possibly the other GO PIs as well
Molecular function overhaul
Paul T would like an update on this.
DOS talked about plan for axiomatization of MF. Requires: 1. Basic axiomatization of simple functions (automated from Rhea) 2. Patterns for axiomatising compound MFs. 3. Logical definitions for new grouping terms suggested by Paul T.
1 & 2 required for 3 to be useful for automated classification
AI: DOS will create GitHub repo, turning discussion on Pauls MF refactoring doc to issues GH site will be used for experiments with MF axiomatisation and inference.
Patterns for compound functions use has_part. Ideally we should be able to use has_part relationship on to generate inferred annotations (see github issue) This plan depends on reliable us of contributes_to in annotation. AI: discuss this inference plan on annotation call.
Becky tested the 'morphogenesis' template. Her questions: 1) We’re now adding in ‘results_in_morphogenesis_of’ relationships (in addition to the cross-products)? 2) Many (but not all) other morphogenesis terms already in GO have ‘is_a: single-organism developmental process ; GO:0044767’ parentage. TG doesn’t currently suggest this parentage, and in most cases it would probably be inherited anyway.
David H tested all templates and suggests that we (Paola) can advertise them as soon as we've added the missing axioms. --Turns out Chris was correct. There were not many that could be added. Most are fly anatomy and fungal anatomy. --So are we good to advertise?
Ready to announce.
protein fibril
Does it belong in GO or PRO? See email thread 'protein fibril' on the ontology mailing list.
DONE: DOS to reply to email thread with 'supramolecular complex' term proposal, also with context from protein complex discussion
HJD - PRO might be better location for disease-caused structures/complexes, for naturally occurring, wouldn't be wrong to add to GO
Adding logical defs to cellular response terms - PUNTED
DOS has started to do this. What's the syntax, and should we look into filling all gaps.
'cell proliferation' terms - PUNTED
Most have logical def, e.g. 'epithelial cell proliferation' is_a 'cell proliferation' that acts_on_population_of 'epithelial cell'. Did we have plans for a TG template? (Not in Jira.)
Note: this is a case where we have a semi-random distribution of terms in the ontology vs annotation extensions. Should we just make terms for all AE usage and retrofit AEs?