Ontology meeting 2012-10-11

From GO Wiki
Jump to: navigation, search

MINUTES: Harold

ATTENDEES:


TermGenie template for non-template-able terms?

At the GOC meeting in Pasadena, the request was made for a new functionality in TG. For new term requests where the current templates cannot be used, some curators would like to be able to submit all information pertaining to the new terms they need via TG (including definition, parentage etc.), and be provided with a temporary GO ID that they may use in Protein2GO immediately (though the annotation would be set to 'pending' while waiting for the TG gatekeeper person to review the new terms). We need to discuss feasibility and strategy.

Minutes by Harold

Use for cases of enzymes, when everything is “easy”; flat.  
How to pass to tracker item if needed.  Note: terms CAN be modified after given an id
Pattern after CHEBI submission tool?
Give ID straight away? 
If force submitters to supply def AND at least one parent, then perhaps # of potential obsoletes will be minimized. (Paola).
But what if the term requests are .. absurd. ?
Jane; only allow via protein2go? ; if not valid only stays in protein2go?		
Could it go into editors version immediately? Or with a tag of some sort.
TermGenie gives immediate email update as to what is going on with the request.
You also get email if approved or not finally.
Entry point: ? Choice box in term genie if no template (create de novo term or 
something like that). 
Review process: Pending tag as being a subset; check or remove
If submit, reviewed; if we decide obsolete, it would appear in the ontology
 but in the  obsolete node; requester gets email to that effect.
Open to everyone or restrict? Don’t advertise? Require SOME sort of training?
(Currently 35 people of the list; most of them are ontology editors now). Open to “advanced annotators”. 

Action: Chris and Heiko to study feasibility. Use examples from current SF items?


Jane: New Single/multi- organism axes.

How to formally define?
OWL to specify cardinality, but on what 
 Has_participant > 1 taxon id?
Single/multi-cellular axes?
 Cardinality of 1 or 2 = single (due to cell-cell signaling0
 Cardinality > 2 = multicellular

Status of Web Protégé II: no reasoning at the moment; argh!