Ontology meeting 2013-11-20: Difference between revisions

From GO Wiki
Jump to navigation Jump to search
No edit summary
 
(10 intermediate revisions by one other user not shown)
Line 1: Line 1:
[[Category:Ontology]]
Attendees: Jane, Becky, Paola, DavidOS, DavidH, Harold, Heiko, Chris, Judy
Attendees: Jane, Becky, Paola, DavidOS, DavidH, Harold, Heiko, Chris, Judy


Minutes: Tanya
Minutes: Tanya


===viral cross-products  ===   
===viral cross-products  ===   


<pre>SET UP SEPARATE CALL - use next Wed SF call time for this </pre>
<pre>SET UP SEPARATE CALL - use next Wed (Nov. 27) SF call time for this </pre>


*We need a better way to represent viral modulation of cell activity. We've used has_particiapnt for now:
*We need a better way to represent viral modulation of cell activity. We've used has_particiapnt for now:
Line 33: Line 33:
has_participant is wrong here...need to say has_participant some protein encoded_by virus
has_participant is wrong here...need to say has_participant some protein encoded_by virus


===Follow-up: TG TEMPLATE FOR ORGANELLE PART (LUMEN, MEMBRANE)===
===Follow-up: TG TEMPLATE FOR ORGANELLE PART (LUMEN, MEMBRANE) ===


See latest discussion:  
See latest discussion:  
Line 41: Line 41:
Do we have action items/time-frames for this?
Do we have action items/time-frames for this?


===Stemming from question of adding 'luteinizing hormone' in GO===
<pre>DavidOS has slides and suggestions.  Keep in mind potential applicability of relations used
for other ontologies.  Compatibility with BFO?  Ex. 'lumen' in GO and 'lumen' in BFO.
DavidOS is on the job.  Will sort out.</pre>
 
===Stemming from question of adding 'luteinizing hormone' in GO (push to next meeting)===


See email thread. Chris says: "GO eds - we will want to switch IDs in the production logical definitions before this, to check the inferences. That's another external ontology imports file - not a big deal. Discuss on thursday?"
See email thread. Chris says: "GO eds - we will want to switch IDs in the production logical definitions before this, to check the inferences. That's another external ontology imports file - not a big deal. Discuss on thursday?"


===Review of template requests in Jira===
===Review of template requests in Jira (push to next meeting)===


* https://www.ebi.ac.uk/panda/jira/browse/GO-58 homeostasis xps
* https://www.ebi.ac.uk/panda/jira/browse/GO-58 homeostasis xps
Line 54: Line 58:
* https://www.ebi.ac.uk/panda/jira/browse/GO-204 Protein Complex by activity
* https://www.ebi.ac.uk/panda/jira/browse/GO-204 Protein Complex by activity


===New TG template for 'cell migration'?===
===New TG template for 'cell migration'? (push to next meeting)===


Note that 'cell chemotaxis' is_a 'cell migration', which is fine based on their definitions. However, both 'cell chemotaxis' and 'cell migration' terms have logical defs based on alters_location_of, e.g.
Note that 'cell chemotaxis' is_a 'cell migration', which is fine based on their definitions. However, both 'cell chemotaxis' and 'cell migration' terms have logical defs based on alters_location_of, e.g.
Line 77: Line 81:


* https://www.ebi.ac.uk/panda/jira/browse/GO-223
* https://www.ebi.ac.uk/panda/jira/browse/GO-223
<pre>There's a plan. Editing shouldn't be very different from xp generation within GO. Try making xps
using external ontologies. </pre>




Line 82: Line 89:


Can all editors test the OE beta?
Can all editors test the OE beta?
<pre> People have tested normal editing, no real test of adding xps to external ontologies yet. Also making
regular relationships to external ontologies - example: axon part_of neuron, process has_participant xxx</pre>


* Day-to-day editing. OK? Should be unaffected
* Day-to-day editing. OK? Should be unaffected
* Loading combined file: go_write_test.obo (file with production xps added) in editors directory
* Loading combined file: go_write_test.obo (file with production xps added) in editors directory
** Save filter (advanced save):
** Save filter (advanced save): (see this figure [[File:Screen_Shot_2013-11-19_at_7.01.50_PM.png]]
** ID starts-with "GO:"
** ID starts-with "GO:"
** don't save relations
** click box for properties
** Do a diff of the output
** Do a diff of the output


<pre>  
<pre>  
Make sure you click []allow dangling references or it won't load.  
Make sure you click []allow dangling references or it won't load.  
MAKE SURE YOU SAVE INTO THE SAME DIRECTORY THAT YOU LOADED IT FROM - because zombie apocalypse otherwise.</pre>
MAKE SURE YOU SAVE INTO THE SAME DIRECTORY THAT YOU LOADED FILE FROM - because zombie apocalypse otherwise.</pre>


==== Downstream changes ====
==== Downstream changes (push to next meeting)====


* Reconfigure TG to write all-in-one
* Reconfigure TG to write all-in-one
* Change pipeline, at first just emulate existing pipeline
* Change pipeline, at first just emulate existing pipeline


=== Taxon constraints ===
=== Taxon constraints ===

Latest revision as of 16:02, 1 July 2014

Attendees: Jane, Becky, Paola, DavidOS, DavidH, Harold, Heiko, Chris, Judy

Minutes: Tanya

viral cross-products

SET UP SEPARATE CALL - use next Wed (Nov. 27) SF call time for this 
  • We need a better way to represent viral modulation of cell activity. We've used has_particiapnt for now:
[Term]
id: GO:0039671 ! evasion by virus of host natural killer cell activity
intersection_of: GO:0030683 ! evasion or tolerance by virus of host immune response
intersection_of: has_participant CL:0000623 ! natural killer cell

but what we really need to do is say e.g. natural killer cell activity = some biological process has_participant natural killer cell and this process is the regulation of that activity.

Also CL:0000451 ! dendritic cell term

  • We've got lots of terms that reference a specific host protein, where the virus modulates one or more functions of that protein e.g.
id: GO:0039548
name: suppression by virus of host IRF3 activity

IRF3 has multiple activities so we wouldn't make an MF term 'IRF3 activity' but we need it somewhere to be able to say the virus modulates it.

  • Processes where the differentia is that it involves a virus e.g.
id: GO:0039699
name: viral mRNA cap methylation

has_participant is wrong here...need to say has_participant some protein encoded_by virus

Follow-up: TG TEMPLATE FOR ORGANELLE PART (LUMEN, MEMBRANE)

See latest discussion:

https://www.ebi.ac.uk/panda/jira/browse/GO-185

Do we have action items/time-frames for this?

DavidOS has slides and suggestions.  Keep in mind potential applicability of relations used 
for other ontologies.  Compatibility with BFO?  Ex. 'lumen' in GO and 'lumen' in BFO.
DavidOS is on the job.  Will sort out.

Stemming from question of adding 'luteinizing hormone' in GO (push to next meeting)

See email thread. Chris says: "GO eds - we will want to switch IDs in the production logical definitions before this, to check the inferences. That's another external ontology imports file - not a big deal. Discuss on thursday?"

Review of template requests in Jira (push to next meeting)

New TG template for 'cell migration'? (push to next meeting)

Note that 'cell chemotaxis' is_a 'cell migration', which is fine based on their definitions. However, both 'cell chemotaxis' and 'cell migration' terms have logical defs based on alters_location_of, e.g.

[Term]

id: GO:0072676 ! lymphocyte migration

intersection_of: GO:0016477 ! cell migration

intersection_of: alters_location_of CL:0000542 ! lymphocyte

Chris says: "It would be good to use a more generic relation, but this depends on the hierarchy being disjoint with other classes using the same relation with a different genus.."


Review proposed fixes to terms for extracellular & extra-organismal region terms - including status of 'other organism'

See: https://www.ebi.ac.uk/panda/jira/browse/GO-216


Moving logical defs into the editors obo file

There's a plan. Editing shouldn't be very different from xp generation within GO. Try making xps 
using external ontologies. 


Testing of OE catalogs

Can all editors test the OE beta?

 People have tested normal editing, no real test of adding xps to external ontologies yet. Also making
regular relationships to external ontologies - example: axon part_of neuron, process has_participant xxx
  • Day-to-day editing. OK? Should be unaffected
  • Loading combined file: go_write_test.obo (file with production xps added) in editors directory
    • Save filter (advanced save): (see this figure
    • ID starts-with "GO:"
    • click box for properties
    • Do a diff of the output
 
Make sure you click []allow dangling references or it won't load. 
MAKE SURE YOU SAVE INTO THE SAME DIRECTORY THAT YOU LOADED FILE FROM - because zombie apocalypse otherwise.

Downstream changes (push to next meeting)

  • Reconfigure TG to write all-in-one
  • Change pipeline, at first just emulate existing pipeline

Taxon constraints

Taxon constraints will be easier to edit when the mega file exists, logical defs go in first, then tax constraints.  
Figure out a way to track publications (PMIDs) for taxon constraints.  OE can't do this right 
now.  Can add after by hand to OBO file. Could move the PMID to the gloss of the def and then add the
PMID to the def dbxref.

Action potential

Planning to generate GO process term for action potential if no objections. See: https://www.ebi.ac.uk/panda/jira/browse/GO-215

 
DavidH: 'generation of action potential'?  action potential is something you measure
DavidOS: No.  Action potential = series of changes in potential across a membrane
Paola: specific lots of stuff in comment, agrees with DavidOS' proposal
All: Go ahead.