Ontology meeting 2012-03-21

From GO Wiki
Jump to: navigation, search

30 minute call

MINUTES: PAOLA


  • UPDATE ON CHEBI SWITCH-OVER

Looking to fully automate creation of chemical metabolism terms, through TG1 templates addition.


  • ESTABLISHING SOP FOR CHECKING JENKINS

Chris, could you please have a look at the link below, and let us know if we're doing things right? Any comments/additions? Thanks.

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


  • MINUTES


Present:

Jane

Becky

Paola

David

Tanya

Heiko

Karen


  • UPDATE ON CHEBI SWITCH-OVER:


Chris' update on CHEBI:

"There is an ontology file called uchebi.{obo,owl} in ontology/extensions on svn. This is chebi with new IDs with the IDspace "UCHEBI" for the union of bases and conjugates. There are alt_ids to the specific CHEBI terms, but these are not in the file.

The ontology x-chebi in the same directory is made automatically from bp_xp_chebi.obo - it uses the alt_ids in the above to map equivalence axioms to the UCHEBI term where appropriate.

You can load x-chebi-import in protege.

I believe TG should be able to use this. However, the challenge is in writing the equivalence axioms out to the bridge ontology."

So we're not quite there yet, but things are moving forward.

For the time being, GO editors should keep hand-editing the goche.obo and go_xp_chebi.obo files in the ontology/editors directory on SVN. Chris and Heiko will let us know when we can stop doing so.

AI: Put the GO-CHEBI alignment on every editors' call agenda until it's done!

How are we going to then maintain the alignment? We need to discuss this with Chris and Heiko. It's on the agenda for next week's call.


  • ESTABLISHING SOP FOR CHECKING JENKINS:


Heiko: the previous build failed because of purely technical reasons.

In general, the quickest way to figure out why a build fails is to look at console output, regardless of the reasoner used. The causes of the failure can be found at the very bottom of the message. If not, we can use Protege.

AI: Becky will update the SOP wiki (http://wiki.geneontology.org/index.php/Jenkins) (done)


  • ADDITIONAL DISCUSSION ON:


  • LOGICAL DEFINITIONS:


Now that all terms in GO are defined, Tanya suggests generating stats with a percentage indicating how many terms have logical definitions.

AI: GO editors should add logical definitions whenever we run into terms that are missing them. Maybe we could take turns to periodically look at the ontology and add a bunch of them in. David and Jane will add this to the list of tasks to give to the PIs.


  • SYNONYMS FOR THE "REGULATION BY REGULATION" TERMS:


Now that the regulation_by template is online on TG1, Becky asks if we want to keep all the default synonyms for these terms or not. Creating a new test term:

name: positive regulation of inositol biosynthetic process by regulation of transcription from RNA polymerase II promoter

generates >100 synonyms in TG.

For 'positive regulation by positive regulation', this number is going to be even higher.

We agree we can probably do without the synonyms, on the ground that these terms are quite specific anyway. Also some of the automatically generated synonyms may have little meaning or sound incorrect, and it would be a misuse of the TG gatekeepers' time to have to go through a list of hundreds of synonyms just to strip the bad ones off. However, if the process in the first bit of the regulation_by term has a synonym, we may want to keep that for users' convenience. We don't need to add synonyms for "reg. of transcription" bit. To save space, we agreed to avoid the usual "upregulation" and "downregulation" synonyms, and stick to pos. reg. and neg. reg. respectively.

AI: Paola to suggest a template, and email round to make sure we all agree.


My suggestions are below. Unless there are objections, I propose we implement them, and follow-up if needed after the regulation_by template has been in use for a few days at least. We should be able to realize fairly quickly if we're still getting too many synonyms (or if we're missing the ones we'd really need).

I'll refer to the same syntax in place in the TG template:

Hint: Example: regulation of signal transduction by receptor internalization

[results in] : regulation of signal transduction [** MUST BE A REGULATION TERM **]

[regulator] : receptor internalization


If the [results in] process has synonyms, add them.

E.g. (SF 3483296) "regulation of thiamine biosynthetic process by regulation of transcription from RNA polymerase II promoter" should have the exact synonyms "regulation of thiamine anabolism/biosynthesis/formation/synthesis by regulation of transcription from RNA polymerase II promoter".


If the [results in] process is a "positive regulation of x" term, do NOT add standard synonyms such as "up regulation", "up-regulation", "upregulation", "activation". But if x has synonyms, add exact synonyms starting with "positive regulation" only (i.e. refer to case above).

E.g. in Becky's example: "positive regulation of inositol biosynthetic process by regulation of transcription from RNA polymerase II promoter" the term should have only the exact synonyms "positive regulation of inositol anabolism/biosynthesis/formation/synthesis by regulation of transcription from RNA polymerase II promoter".


If the [results in] process is a "negative regulation of..." term, do NOT add standard synonyms such as "down regulation", "down-regulation", "downregulation", "inhibition". But if x has synonyms, add exact synonyms starting with "negative regulation" only (i.e. refer to case above).

E.g. (SF 3471534) "negative regulation of cellular hyperosmotic salinity response by negative regulation of transcription from RNA Polymerase II promoter" should only have the exact synonym "negative regulation of cellular response to hyperosmotic salt stress by negative regulation of transcription from RNA Polymerase II promoter".


If the [regulator] process has a synonym, use it.


If the [regulator] process is regulation of transcription from RNA polymerase II promoter or positive regulation of transcription from RNA polymerase II promoter or negative regulation of transcription from RNA polymerase II promoter, do NOT use synonyms for the regulator process.



  • ONTOLOGY CHANGE REPORTS:


Any news on where the ontology change report emails have gone?

AI: Chris says: When Seth is back on Thursday, we'll log into the change reports machine and check the crons to see what's going on. (Heiko adds that it should be possible to have Jenkins handle these reports.)


  • COLUMN 16:


At the editors' call on April 18th, we'll have a dedicated meeting with Val and Midori to co-ordinate the relationships being in column 16 and the ontology.