GO-CAM May 24th, 2017: Difference between revisions

From GO Wiki
Jump to navigation Jump to search
 
(20 intermediate revisions by the same user not shown)
Line 13: Line 13:


= Minutes =  
= Minutes =  
*On call: Chris, Karen, Kimberly, Moni, Pascale, Seth, Suzi
*On call: Chris, Karen, Kimberly, Moni, Pascale, Paul T., Seth, Suzi


== Software Updates ==
== Software Updates ==
Line 25: Line 25:
**Curators can now annotate with the reasoner on and this should catch mistakes
**Curators can now annotate with the reasoner on and this should catch mistakes
**Ultimately working towards real-time error checking
**Ultimately working towards real-time error checking
**If using undone/redo DON'T USE THE REASONER
**If using undone/redo DON'T USE THE REASONER right now - still some troubleshooting to be done here
**Removed GAF output option from Noctua
**Removed GAF output option from Noctua
*Human readable GPAD output
*Human readable GPAD output
Line 31: Line 31:
**See Annotation preview under Workbenches menu
**See Annotation preview under Workbenches menu
**Will live update as curators edit their models
**Will live update as curators edit their models
*Added visual compactor
**Individuals will map into approx. 700x700 square, essentially a screen view
**This helps keep individuals closer together
**One caveat - individuals can get stacked
**Can individuals connected together get stacked?
***There are a number of possible variations to this display - can talk more about the options here
**Exploding specific box(es) would be useful; there is a ticket for this
*Ability to have different views of annotation sets
**Annotations from a specific paper
**Annotations for a specific gene (this might need some filtering options)
*Model searching - in Noctua
**Search by gene, GO term, paper
*Browse function in Noctua
**An alpha version of faceted searching in AmiGO
== Direction for these Calls ==
*Use these calls to prioritize
**High priority vs showstoppers - what are the differences?
*What do we need to do to get everyone using Noctua?
*Handling evidence properly - high priority
*Generating a tabular interface - high priority
*Curators should evaluate their workflow for areas where we can improve efficiency - high priority
*Efficiency of adding evidence - not high priority, but would improve workflow
*Import existing CC and BP annotations - high priority
**GP buddy works on CC
**There are issues to work out wrt importing annotations and making sure we are consistent with an instance-level representation
*Constraining relations - not a showstopper, but would make workflow easier
*Documentation - high priority
== Other Considerations ==
*How is Noctua modeling changing our annotations?
*What is the best way to assess this?
**Changes to annotations, e.g. more precise MF representation
**Improved representation of causal relations in BPs
**Changes to enrichment analysis


[[Category: LEGO]]
[[Category: LEGO]]

Latest revision as of 12:00, 24 May 2017

Bluejeans URL

https://bluejeans.com/969313231

Agenda

Recent Software Updates

  • New fast reasoner: Arachne
  • Export GPAD menu option now using reasoner (ie Jim's code; Heiko's is retired)
  • Readable GPAD ("labelified-GPAD") progress

Review of github Tickets

Minutes

  • On call: Chris, Karen, Kimberly, Moni, Pascale, Paul T., Seth, Suzi

Software Updates

  • Updated relations available in Noctua (e.g. directly positively regulates)
    • Also updated existing models
  • Incorporated SynGO models files generated by David OS
    • JSON -> OWL
  • New reasoner
    • Incorporated Jim B.'s Arachne reasoner into Noctua
    • Much faster than ELK
    • Curators can now annotate with the reasoner on and this should catch mistakes
    • Ultimately working towards real-time error checking
    • If using undone/redo DON'T USE THE REASONER right now - still some troubleshooting to be done here
    • Removed GAF output option from Noctua
  • Human readable GPAD output
    • First version is available for use
    • See Annotation preview under Workbenches menu
    • Will live update as curators edit their models
  • Added visual compactor
    • Individuals will map into approx. 700x700 square, essentially a screen view
    • This helps keep individuals closer together
    • One caveat - individuals can get stacked
    • Can individuals connected together get stacked?
      • There are a number of possible variations to this display - can talk more about the options here
    • Exploding specific box(es) would be useful; there is a ticket for this
  • Ability to have different views of annotation sets
    • Annotations from a specific paper
    • Annotations for a specific gene (this might need some filtering options)
  • Model searching - in Noctua
    • Search by gene, GO term, paper
  • Browse function in Noctua
    • An alpha version of faceted searching in AmiGO

Direction for these Calls

  • Use these calls to prioritize
    • High priority vs showstoppers - what are the differences?
  • What do we need to do to get everyone using Noctua?
  • Handling evidence properly - high priority
  • Generating a tabular interface - high priority
  • Curators should evaluate their workflow for areas where we can improve efficiency - high priority
  • Efficiency of adding evidence - not high priority, but would improve workflow
  • Import existing CC and BP annotations - high priority
    • GP buddy works on CC
    • There are issues to work out wrt importing annotations and making sure we are consistent with an instance-level representation
  • Constraining relations - not a showstopper, but would make workflow easier
  • Documentation - high priority

Other Considerations

  • How is Noctua modeling changing our annotations?
  • What is the best way to assess this?
    • Changes to annotations, e.g. more precise MF representation
    • Improved representation of causal relations in BPs
    • Changes to enrichment analysis