Ontology meeting 2016-08-18
Regrets: Paola, David H
Minutes: David OS
On June 23rd, we emailed the go-consortium as follows:
"The TermGenie *Free Form* Term Submission tool will be unavailable for some time, due to changes in the GO release pipeline. (The TermGenie *template* tool is not affected.) Therefore, please note the following:
- If you need to request new terms, please check the many available TG templates to see if there’s an appropriate one to use. You may use the search function at http://go.termgenie.org, or navigate the templates by expanding on the list there.
- If no template is available, please open a new ticket on the GitHub tracker at https://github.com/geneontology/go-ontology/issues. Please follow the guidelines to help us process your request faster and better: https://github.com/geneontology/go-ontology/blob/master/CONTRIBUTING.md
We may review the situation later on (not sooner than August)."
August has come and someone has used TG freeform - there are/were 3 terms in the queue this week (which I will not approve; I'll contact the submitter and add them manually). Should we/can we disable TG FF completely, as most editors agreed would be best? Happy to send a reminder to go-consortium as a temporary measure.
FreeForm is currently broken so needs to be closed down. But could still be useful if fixed, e.g. for protein complexes (better for Birgit to add) AI: Chris to post ticket detailing problems and to look into whether there is some easy way to fix.
Does this need to be discussed at an editors call? At the time of writing, consensus seems to be to discuss at an annotation call instead.
- it pertains to the ontology group as we should be thinking about the different distributions of GO and what their purpose is Cjm (talk) 15:24, 18 August 2016 (UTC)
Instantiating annotations inferred from inter-ontology links (F-P etc) - should we stop doing this centrally?
Problem of divergence: Berkely vs EBI. 2 use cases: Display (Pombase); grouping (a problem for resources that use GO-basic, which lacks inter-ontology links)
Who depends on the central job? PomBase (needs it for display) TAIR MGI (Needs it for grouping AND display) First step towards ultimately removing central inference job - adding inter-ontology links to GO-basic.obo
AI: Chris will announce with some clear timeline.
See discussion here https://github.com/geneontology/go-ontology/issues/12578
- for GO the meaningful groupings are morphogenesis vs development etc (which I suspect are not used consistently by annotators anyway). The additional high level distinction of whether something is an organ appendage tissue organ-part etc I feel are not so useful Cjm (talk) 15:25, 18 August 2016 (UTC)
Decision: split Organ development into plant & animcal organs Fungi don't have organs: at least as far as GO is concerned. Push these (e.g. spore forming organ) up under anatomical structure development.
Daily report of terms and stats
Currently, we (editors) receive a daily report of inference changes, but as of July 12th, the wider group (go-watchers) no longer receives the daily report of definition changes, obsoletions, term stats etc. The two reports do not fully overlap in aim and group target. Both are useful. Do changes in the pipeline mean we can't have the watchers email, or did it just slip through and can be easily revived?
- if emails are not going out to go-watchers, this is a bug. Report on go-site. This should have nothing to do with the inferences report, which is separate Cjm (talk) 15:23, 18 August 2016 (UTC)
go-watchers email no-longer going out. Got lost in refactoring pipeline. AI: Chris to see if there is a quick fix, otherwise will file ticket to look into this later.
Review capitalisation of terms
Is there still a plan for a Noctua meeting before or after the GOC? Chris to bring up at next managers call - or directly with Paul. Tanya: we need to know soon in order to drum up attendance.
== Meetings during Noctua training at EBI (Aug 31st-Sept 2nd)
David OS and Chris M to meet Wed afternoon - ask Tony S if he can join to discuss annotation inference pipeline.
Major Aim: Fast inference for LEGO. Investigating RDFox right now. David: Can this also be used for inference on AE - web service? Chris: Should be possible to use Jim's OWLery (EL) server for this. Might be more straightforward.
Chris; Jim has also been known to write Protege plugins. Any priorities for this?