Ontology meeting 2012-08-29

From GO Wiki
Revision as of 15:29, 1 July 2014 by Gail (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

30-minute meeting

MINUTES: Harold

ATTENDEES:

  • Harold
  • Becky
  • Tanya
  • Chris
  • Heiko
  • Judy


CHEBI PAPER UPDATE

How's it looking after Chris' meeting with Tanya and David?

Tanya reporting: a few sections were moved around to sound less methody; David will add a bit more, and 
a couple of tables will be added, one with a X-product example.

RESPONSE_TO TERMS

  • These cross-products are now in the main ontology files. Tanya and David have gone though the no-parse list and cleaned up the ontology. All of the rest of the terms in the no-parse list are either not chemicals or are proteins.
Tanya reporting: example: water not a chemical stimulus but an abiotic stimululs, since the responses are
actually more to pressure than to water as a chemical. "response to" any proteins: CHEBI will not have proteins
so won't get X-products yet.
 
  • What's the time-frame for a 'response to' template in TG1?
X-products are now in GO-XP-CHEBI, so Heiko can proceed; will use Midori's SF items to test (within hours! this is done and tested.. 'cellular response to' not yet ready for template form).


PROTEIN LOCALIZATION_TO TEMPLATE

  • The localization terms in GO have been standardized to 'localization to' (except the maintenance ones, which have remained as 'maintenance of localization in'.
  • Can the TG template be created now based on Chris' SOP? [1]
Chris will do a final parse to add missing logical defs to TG template, then can proceed. 
relationships: same relatiohsip used in CHEBI for transport.

Adding asserted implied links to gene_ontology_write.obo

Where we're up to, and what to do next?

CJM (via email with subject: adding inserted links to gene_ontology_write.obo):

  • I think all will be well if we only assert/retract subclass axioms only. That way you don't get equivalence axioms materialized as relationships.
  • I think the easiest way is an option in Oort that masks this step. Or in fact Heiko if we just to do this directly as an owltools command (DavidOS has a tracker item for this anyway). E.g. "--assert-inferred-subclass-axioms" with a sub-option for removing redundant ones.
  • We have to get away from asserting these in the editors file though. The options are abandone oboedit or allow Elk to be called from within OE.


OORT contains now a standalone tool to assert inferences: obo-assert-inferences
Usage: obo-assert-inferences editors/gene_ontology_write.obo editors/go_xp_chebi.obo extensions/bio-chebi.owl

This will assert the inferences in gene_ontology_write.obo, to write it into a different file first, specify an output file with the -o

Becky will try Heiko's script; this will add the inferences made for the chemical terms.
Lego plugin demo on Sept 4th: would be good if all Cam-editors can attend. 
Tanya and David have experience in adding LEGO examples into Protege, so can bug them for their expertise in the meantime.