Ontology meeting 2013-08-08

From GO Wiki
Jump to navigation Jump to search

Minutes: Paola

Attendees: Paola, Chris, David, Harold, Heiko, Judy, Becky


Protege

Judy: Protege is not funded. Idea: write to Mark Musen and see what his plans are. Otherwise it seems we'll have to look for other ways? Judy could act as connector person. Chris suggests bringing in someone else, Suzi and Helen and David OS (and the Manchester group possibly) and see what the best course of action would be.

AI: Chris will get the ball rolling before he leaves for his holidays.


How do annotations propagate up has_part links

Stemming from this SF ticket from Birgit: https://sourceforge.net/p/geneontology/ontology-requests/10289/

See the current existing links for:

  • GO:1990156 DnaB-DnaG complex has_part GO:1990161 DnaB helicase complex

and

  • GO:1990161 DnaB helicase complex is_a GO:0033202 DNA helicase complex

QUESTION: The DnaB-DnaG complex is capable of helicase activity. Will it break anything if we also add-in:

  • DnaB-DnaG complex is_a: DNA helicase complex

(David) I just tried making DNAB-DNAG complex is_a DNA helicase complex in Protege. Running both ELK and Hermit didn't seem to throw any problems.

AI: David and Harold will do a final pass on the file (because the comments will then be lost), then pass it on to Chris. Then Chris will give a two-weeks heads-up before moving the capable_of in the main GO.

In the meantime, we can still add the capable_of in the xp file for Birgit's terms.


Transport vs. transmembrane transport

Stems from https://sourceforge.net/p/geneontology/ontology-requests/10288/

Underlying policy question: If we determine that a class only consists of a specific subtype, is it ok to simply redefine that class, or should we still create a new subtype. In this case, if we have a term 'X transport' in the ontology, but the only type of 'X transport' is 'X transmembrane transport', can we simple reclassify 'X transport'? How do we specify explicitly that ALL 'X transport' is in fact 'X transmembrane transport'?

AI: David will create new 'x transmembrane transporter' terms, and will give Val the bad news that she'll have to move her annotations. That'd be the safer option so we don't risk to break anything. David will also check the function-process links, but it's likely that Val's slim is not using MF.


RHEA MAPPINGS

To try and get a 1:1 mapping for GO terms and Rhea Xrefs, I've (Becky) started going through the list of

  1. GO terms with multiple Rhea X-refs (36)
  2. GO terms that share Rhea X-refs (69 GO terms)


RHEA MAPPINGS FOR MULTISTEP REACTIONS

For the multiple X-refs, some are where the reaction is multistep, and Rhea has IDs for the overall reaction PLUS the individual steps. Do we just want to keep the Rhea Xref for the overall reaction, or all of them?

E.g. the three-step reaction 9beta-pimara-7,15-diene oxidase activity ; GO:0036209 (1.14.13.144)

OVERALL REACTION: 9β-pimara-7,15-diene + 2 H+ + 3 NADPH + 3 O2 <=> 9β-pimara-7,15-dien-19-oate + 4 H2O + 3 NADP+ (RHEA:31954)

Individual steps:

  1. 9β-pimara-7,15-diene + H+ + NADPH + O2 <=> 9β-pimara-7,15-dien-19-ol + H2O + NADP+ (RHEA:31958)
  2. 9β-pimara-7,15-dien-19-ol + H+ + NADPH + O2 <=> 9β-pimara-7,15-dien-19-al + 2 H2O + NADP+ (RHEA:31962)
  3. 9β-pimara-7,15-dien-19-al + NADPH + O2 <=> 9β-pimara-7,15-dien-19-oate + H2O + NADP+ (RHEA:31966)

NB: The same will apply to KEGG Xrefs in many cases. GO:0036209 has 4 KEGG Xrefs.

AI: We will use the top (overall) rhea reaction.

RHEA MAPPINGS FOR NAD(P)H

In some cases, we have multiple Rhea mappings for NADPH vs NADH. E.g. 

3-hydroxyindolin-2-one monooxygenase activity ; GO:0036192
Catalysis of the reaction: 3-hydroxyindolin-2-one + NAD(P)H + H+ + O2 = 2-hydroxy-2H-1,4-benzoxazin-3(4H)-one + NAD(P)+ + H2O.
EC:1.14.13.139

Do we want to split the GO term into two, or keep both RHEA xrefs?

AI: We will use the generic GO parent and keep the two RHEA xrefs. Becky will ask Rhea if they can create superclasses in those cases where there are two RHEA xrefs, so the GO term can refer to that (have emailed Rhea via their web contact form).

RHEA MAPPINGS FOR (d)CMP

E.g. cytidylate kinase activity
Catalysis of the reaction: ATP + (d)CMP = ADP + (d)CDP.
  • EC:2.7.4.14
  • RHEA: 25097
  • RHEA: 11603

AI: Becky will make 2 separate GO terms because ChEBI doesn't seem to have a 'superclass', but she'll check with Rhea if there's any reason why not.


GOC MEETING IN BAR HARBOR

Judy says that the meeting will officially be over by Saturday. If anyone wishes to suggest topics for break-out sessions before or after the main meeting, it might be good to start planning, so rooms can be booked accordingly.

AI: E.g. the GO editors might want to meet with Peter D'E. on Sunday morning - Becky will email Peter D'E to check his travel plans (have emailed Peter)?

Retrieved from "https://wiki.geneontology.org/index.php?title=Ontology_meeting_2013-08-08&oldid=52926"