Ontology meeting 2013-01-24

From GO Wiki
Jump to navigation Jump to search

MINUTES: Jane

ATTENDEES:

ChEBI paper

Anything missing/left to do before (pre)submission?

-David is putting in the presubmission today

Annotation group

Request to discuss function-process links and annotation to processes connected by MF-BP links. Next week ok?

-Yes, okay

'Releasing' TermGenie free form

Jane has prepared a list of candidate users. Next step: circulate/advertise. DONE. TermGenie gatekeepers, please remember to delete the termgenie_unvetted subset line if a new term looks straightforward. Suggested strategy for non-straightforward terms: contact submitter; if necessary, discuss among editors; any more comments/suggestions?

-Looks good, we'll try this and see how it goes

Sync issue between editors version and release versions?

We still haven't got to the bottom of this. The example is GO:0019064. In go.owl, go.obo etc it has three parents. In go-trunk/ontology/editors/gene_ontology_write.obo it has two (missing GO:0048610 cellular process involved in reproduction which I removed).

-This relationship is somehow being inferred, let's look into it next week

Protein/glycoprotein/lipoprotein terms in ChEBI

SF request on ChEBI tracker: https://sourceforge.net/tracker/?func=detail&atid=703818&aid=3592472&group_id=125463

There was an email thread ('protein in CHEBI and PRO') between Chris and Darren, but no conclusion was reached as far as I [Paola] can see. Chris, was there any further follow-up to your message below:

'CHEBI is open ended here:

CHEBI:33709 ! amino acid [DEF: "A carboxylic acid containing one or more amino groups."]

It doesn't *exclude* the amino acid including attached bits.

It seems the isa/has_part confusion permeates CHEBI even after all our work aligning with GO. Or maybe it's just me that's confused - is there some documentation I'm missing?

Anyway, it seems the direction we're going is a more liberal definition that is inclusive of attached bits. Does this fit with our intended inferences in GO?'

-Chris following up

Merging vs. obsoleting cellular component terms: chromatin assembly complex AND chromatin remodeling complex

see LONG SF discussion here: chromatin assembly/remodeling complex terms

When the terms were created, the complexes studied were shown to be involved in separate processes. More recent experiments have shown that some chromatin remodeling complexes can also assemble chromatin. The current proposal is to merge the two terms into "chromatin assembly or remodeling complex." None of us like the 'or' in the term but we can't think of a better name. Obsoleting the terms has big repercussions wrt annotation review and changes to mapping files and subsets. Merging would mean that annotations would still be correct and would not have to be reviewed.

OK to merge?

-Chris suggested making this sort of structure:

chromatin assembly complex
---[i] chromatin assembly and remodeling complex ; GO:new

chromatin remodeling complex
---[i] chromatin assembly and remodeling complex ; GO:new

Daily digest emails for new/edited/obsoleted terms (obo-diff.html)

Heiko/Chris, is it feasible to 'convert' the Jenkins diff into emails (what curators really wish for)?

 If we want to do this e-mail send via jenkins, there is an issue with the custom e-mail plugin.
Heiko, Seth and Chris to work on this

Ontology Workshop Jan 2013

http://wiki.geneontology.org/index.php/Ontology_workshop_Jan_2013

  • Agenda
  • Logistics

Update on cell cycle meeting

Dates confirmed: Feb. 28th and March 1st.

Confirmed attendees: Val, Jacqueline Hayles, David, Tanya, Jane, Paola, Bijay Jassal (Reactome/EBI), Lisa Matthews (Reactome/New York), Mary Dolan (MGI - attending remotely).

TBC: Takashi Toda? Rob De Bruin? Peter Fantes?

Agenda: to be prepared. Arrange conference call after the Berkeley workshop?

Ontology files

We really need to do something to improve our file downloads from the website. I couldn't figure out how to get go.owl from the website yesterday! Do we have a plan for transitioning to the new, simplified file structure? Can we make a plan of action please?