Ontology meeting 2014-05-08

From GO Wiki
Jump to navigation Jump to search

Attendees: David OS, Paola, Chris, David H, Harold, Heiko, Judy, Tanya, Jane, Paul T

Minutes: Paola

Ascii-check

It's easy to introduce non-ascii characters when hand-editing TG terms before commit.

AI: Heiko will implement an ascii-check in TG.

Ontology requests backlog in SF

Discussion on possible strategies and resource allocation. Being on top of ontology development issues in general is a priority for the GOC, and ontology requests are a relevant part (see minutes from the last Managers call, http://wiki.geneontology.org/index.php/Manager_7May2014). At the first quarterly GOC call (probably June 18th) we'll all need to report on priorities, and choose targets for the following quarter. Roll-over funding should be available soon, and we could use it for focused meeting. A SF jamboree may be organized by video-conference, to solve tickets that require global discussion (possibly with people from the West Coast traveling to the East Coast to reduce time difference with UK). But, given funding, we may want to meet to work on other ontology matters too. OTOH, a jamboree may not be the best solution to work on ontology requests that simply require time; several are mini-projects in themselves.

Note that the new Sourceforge API (Allura) makes it very easy to generate text file reports from any search for tickets that you can do using the SourceForge web interface, (such as this one. Reports can include any content of tickets returned by search.

AI: Ask David OS if you have any requests for reports or report generating scripts that might be useful for triaging SF ticket backlog.

AI: Editors to devote 20% of weekly time, every week, to open ontology requests, until the first quarterly GOC call on June 18. There, we'll report on progress and see if we've made a dent. Editors on weekly rotas are likely to spend ~50% of their time on SF and TG review alone. As of May 9th, there are 283 open SF tickets.

AI: tag difficult SF tickets to highlight that they are projects (big or small) rather than simple requests. Tanya created a label called 'mini-project', please apply it to your relevant tickets so we have an idea of numbers and time needed.


Links to annotation guidance

Stemming from go-discuss thread "GO:0001816 ! cytokine production and it's children"

Shall we add 'notes' to GONUTS manually, or shall we seek a more automated strategy for displaying links to annotation guidance directly to GO terms and their children, maybe as a new field? E.g. 'annotation_guidance_link'

The aim is for annotation guidance to be visible regardless of GO browser used.

Chris suggests to do this solely in Protege by a new annotation property, called e.g. 'annotation_guidance_link'; we'll also be able to indicate if the link should apply to all descendant terms.

AI: Chris will put the new annotation property in, then we can all use it.

Editing in Protege

Tanya sent guidelines on May 5th. Have we all tried them out?

AI for all: Tanya posted the guidelines here: http://wiki.geneontology.org/index.php/Editor_Guide#Editing_in_Protege. By the next meeting, we will all test and report. To start with, we should try the roundtrip and make sure we get nothing in the diff.

Cell motility design pattern

See email thread with same subject. Stemming from https://sourceforge.net/p/geneontology/ontology-requests/10719/

AI: Discuss next week

Multicellular org transport

lipid transport across blood brain barrier - I think this needs a parent 'multicellular organism transport' which we don't have. Any objections to my adding it?

David H suggests to go more granular instead, and link terms of this kind to e.g. existing 'transepithelial transport'. No need to add 'multicellular organism transport' for now; when we have UBERON in the editors file, we can probably define better.

Property chains

Once we all start working in Protege in earnest we are going to start adding property chains for completeness of reasoning. For example we will want things like 'starts with o 'has input' SubPropertyof 'has input'. Is there a way to add these and have them in the OBO file?

  • import issue should now be fixed
  • the protocol is to add property chains in RO. Any relation axioms added when editing GO will be lost. Contact cjm or dos for write access, ID ranges, etc. Post ticket on RO google code site to allow discussion before adding new relations/axioms.

We will push this to a later call as necessary.

Inferring parts

What needs to happen for us to be able to do this?

Keep this in mind and discuss at a later call (maybe next week).

Retrieved from "https://wiki.geneontology.org/index.php?title=Ontology_meeting_2014-05-08&oldid=51223"