Ontology meeting 2012-09-20
From GO Public
ATTENDEES: Becky, Paola, Heiko, Harold, Brenley, Tanya, David, Jane, Chris, Judy
Follow-up: ChEBI paper
Heiko has done a pass. Chris has started another pass. David also to go through and add comments. Jane is keeping track of all this in Jira.
Follow-up on homeostasis xps
For background, see http://wiki.geneontology.org/index.php/Ontology_meeting_2012-09-12#Homeostasis_xps and email thread
Chris says: let's get the synonyms done first, then we can do another parse. Maybe a compromise is to tag the synonyms somehow, so we don't have to use the ion name in the label. Chris would put the 'systematic synonyms' in GO.
AI: David and Tanya have prepared a list; Chris will automate it.
Follow-up on localization template
Agreed that 'transports_or_maintains_localization_of' is generic enough to cover both localization and transport. We should replace the existing results_in_localization_of in 'biological_process_xp_cellular_component.obo' with the more generic relation.
- Do any other files need updating after that? What needs to be done next?
- For 'targeting', do we want to use 'results_in_transport_towards' or 'results_in_transport_to'?
- There's a SF item as a test for a new TG template: https://sourceforge.net/tracker/?func=detail&aid=3556059&group_id=36855&atid=440764
Discussion: we agree that for 'targeting', we want to use 'results_in_transport_to'.
AI: Targeting, establishment: Chris suggests to make an ontology that has is_a links to transport stripped out, and then look at the reasoner report. Chris suggests one person does this as a sort of quick analysis. Becky will make a start at it and then hand it over (to David) in case it's not finished before she goes on leave.
Filling in missing xps
As we've added new sets and templates we've missed loads of xps - we should go back and fill them in. We could perhaps divide them up between the editors?
Discussion: The easy ones are very quick, but many are not straightforward.
David and Tanya will work on the involved_in (they had already started).
A Jira issue will be made. AI: Jane to keep track. The Jira issue could also be used to record an SOP. Will be useful for quarterly reports.
Protein-specific secretion terms
We've had a request to add secretion terms that are specific to proteins. GO already has many (mainly cytokine) named protein secretion terms, and this new set are important for disease. Should we add them in?
Discussion: In theory we could use the Protein Ontology to do this, but in reality that's quite flat... We should make our "own" protein ontology and use it to give more structure to such terms.
The submitter should use column 16, but in this submitter's case it is not implemented. We should encourage him to use it, in which case Becky could speak to Rachael to see if he could use Protein2GO.
However, it looks like he needs the terms for ontology building so Chris is open to dialogue: Can they use OWL class expressions? (Becky will ask).
Placement of "secretion" is also discussed. It involves directionality, as a kind of export.
We only want to map to the exact Rhea reactions, otherwise we'd get redundancies.
Discussion about acknowledging people on the GO website. Stemming from wanting to acknowledge people working on the ELK reasoner. Likely applicable to other softwares we might be using. Chris will add something on the GO pages.