Manager 3Nov10

From GO Wiki
Jump to: navigation, search

Agenda

AmiGO testing

(Rama) AmiGO version 1.8 went out for testing 2 weeks back and not many GOC curators came forward to test (some of you may have noticed that Val tested the interface a lot). It is disappointing. And it is not like there is a new feature in AmiGO every week. Here is a breakdown of the groups being supported by the GOC. Can we expect some % of time from each of these groups to test?

  • EBI - GO-Office
  • SGD/Stanford
  • MGI/ Jackson Laboratory
  • EBI - GOA
  • WormBase/CalTech
  • TAIR/ Carnegie
  • Dictybase/Northwestern
  • RGD/Medical College of Wisconsin

Several people seconded the proposal. Should expect GOC-funded people to contribute to the project. Need to provide a testing script for each release of things to be tested. An automated script tests a lot of things, when moved to Stanford servers Rama tests again then sends out email to ask for more testing. Suzi recommends that for the penultimate testing, include a checklist of things to test, also there should be certain groups that are obligated to test at a minimum the new features. Should combine the two groups from EBI and remove RGD from the list. Updates are about every 6 months. Need to make a focussed effort to get the appropriate people to test. Also to get some external users of AmiGO to test. Could send out an email on gofriends or mine the gohelp messages for users. Can also ask the Plant Ontology people as this is their main browser.

Infrastructure

  • Recommendations for handling goa PDB files, to be sent to the GOA group.

Filtering - PDB file should not be filtered on the GOC site. Chris will deal with this. Resolvable URL - need to contact Helen to get resolvable URL for the chain IDs. Rachael to contact Helen Berman and PDB helpdesk to ask about this. Rachael to update Chris on other matters in his email.

Reference Genome

  • PAINT curation ramp up/testing of curation rates: curators: Mike L, Rama, Dmitri, Li?
  • Plan:
    • PAINT bug fixes, curator training (Nov)
    • 10 days in December from each curator

Not two weeks solid but just a couple of hours a day. Keep track of number of hours to do stats on the curation rates. Training on PAINT should only take a total of one to two days.

  • new directory for derived pre-submission PAINT files

If all agreed on directory structure then Chris will ask Mike to add. Do MODs have to promote the PAINT files to a directory for release? No, MODs could slurp the annotations into their files and decide whether to check them or not and then release them with their usual file release. Want the files to be available as soon as possible. Suzi and Chris to deal with this.

  • Ref Genome Software Development: additional projects for Princeton?

Sven still working on annotation tracker, if there are other projects appropriate for Sven please suggest any. Judy suggests whole set of families that have one or two members that have experimental data, promote annotation of these families. Paul Thomas is working on this but Sven would be happy to help out. Software people should have a call to discuss potential projects.

Ontology Development

  • Signaling meeting proposal for February 2011

Proposed budget from Becky seems reasonable, will forward the proposal to GO-tops. Meeting will be in UK, including Ruth, Val, Varsha, Midori, Mike Tyers, UCL, Sandra Orchard, Peter D'Eustachio and Alex Diehl. Mike Tyers knows a lot about signaling. Rachael will also attend to give additional curator point of view.

  • Do we need a face-2-face GOCHE meeting in February?

Plan is to have GOCHE exist as we would like to see it in GO, align with ChEBI unless we think ChEBI is wrong. Midori is working on Amino Acids and their derivatives and they don't align well with ChEBI. Would be good to put in a budget proposal for this, need to spend the funds before end of February. Harold and David/GOCHE people to write a proposal and send it to GO-tops.

  • Staffing at EBI - Jane wants to discuss this. Will discuss this on a GO-top call.
  • Neural-Immune GO in BioPortal - paper published creating a modified subset of GO, picked up by BioPortal, looks like separate ontology but is actually a subset of GO and no attribution to GO. GOC Terms of use are contravened, these are OK as they are and don't need any amendment. Alex agreed it should not be posted as a maintained GO slim. Jane to email BioPortal support desk and the authors of the paper highlighting the problems.