LEGO May 18, 2015

From GO Wiki
Jump to navigation Jump to search

Action Items from May 4, 2015 Call

  • Make a github ticket for adding IntAct protein complexes to Noctua. [DONE]
  • Heiko will let Seth know that curators may add feature requests, but these are not immediate needs.
  • David, Kimberly, and Rama will review their March paper models to compare with Paul's version.
  • Continue to work on LEGO model for the April consistency exercise -

March Paper - Modeling the Mitotic Exit Network (Ras/Hippo kinase-based signaling pathways)

  • Additional reading: The Mitotic Exit Network: new turns on old pathways. and references cited therein. Also: Asymmetry of the budding yeast Tem1 GTPase at spindle poles is required for spindle positioning but not for mitotic exit.
  • Steps/functions to model:
    • Tem1 and Cdc15 physically interact in a manner dependent on Tem1's GTP-binding domain; this interaction recruits Cdc15 to the daughter-bound spindle pole body
      • Tem1 - Cdc15 MF: protein binding IPI (Asakawa, 2001)
      • Tem1 - MF: protein binding involved in protein localization to daughter spindle pole body during late anaphase (term doesn't exist, but that's perhaps the idea to capture) IPI with Cdc15
    • Cdc15 phosphorylates Dbf2 only when Dbf2 is present with Mob1
    • Dbf2 phosphorylates Kar9
    • Cdc14 dephosphorylates Kar9
    • Kar9 localizes asymmetrically to spindle pole bodies and interacts with the Type V myosin Myo2
    • Bfa1 physically interacts with Bub2 and with Tem1 (different papers)
    • Bub2 executes the GAP activity - evidence for this?
    • Bfa1/Bub2 complex positively regulates Tem1 GTPase activity, resulting in negative regulation of Tem1 GTP binding and thus negative regulation of Tem1 interaction with Cdc15
    • Lte1 is required for localization of Bfa1 to the daughter-bound spindle pole body
  • Questions
    • How would we represent asymmetric spindle pole body localization, i.e. mother- vs daughter-destined spindle pole body?
    • Tem1 has GTPase activity, but its active form in this pathway is as a GTP-bound protein. How could we represent this? Use a PRO ID for 'GTP-bound Tem1' and use that to annotate to Cdc15 binding and regulation of protein localization to spindle pole body?
  • Other organisms
    • C. elegans does not appear to have Tem1, Bfa1, or Cdc15
    • Closest C. elegans Bub2 is TBCK-1, but that appears to only be in the TBC domain, and there is no data about this protein
    • Dbf2 and Mbo1 homologues?
    • C. elegans does have a Cdc14 phosphatase


  • With respect to the yeast paper on GTPase, does it make sense to curate from a single paper? If a given paper tells the entire story, then it is fine. But if it indicates that there is more to that story then we should take the time to build a complete model. The yeast paper on Tem1/Bfa1/Bub2 was a pure biochemsitry paper and we should be looking at other papers to say the whole story
  • There was no evidence in that paper on protein binding, bridging activity (protein adaptor). It is more of author speculation. We have been moving away from TAS/NAS. We should apply to same rules here. Curator should capture it only if there is evidence from other papers.
  • GAP activity: GAP proteins activate the GTPase, i.e. GTPase hydrolyses the GTP and now the GTPase is bound to GDP which is the inactive form. GTP bound form is the active form. When Tem1 is bound to GDP it is inactive and thus this negatively regulates mitotic exit. How do we capture the fact that GAP activates the GTPase activity but it makes the GTPase inactive in this process? We should be able to represent GTP bound and GDP bound forms in LEGO? Can we do that using PRO? We can check with Harold.
  • Feature requests:
    • We should be able to edit any part of an annaton (submitted a github ticket)
    • the tool seems to be down a lot. Can we get a schedule from Berkeley on when the tool is brought down for maintenance/data loading?
    • interface shouldn't save until we say save (submitted a github ticket)
  • How about a hackathon-annotation-thon for LEGO? We are not there yet. We should build more models to see what other features are needed etc.