Ontology meeting 2018-09-10
Conference Line
Agenda
Editors Discussion
Order of Operations for External identifiers
For CHEBI ids, new ECs, Rhea? Should GO editors or requestors follow up with EC and/or RHEA requests?
Can we put terms into GO while waiting for ids outside of ontology?
Project Update
- RO meeting to take place in Colorado in October
Testing the new version of Protege
There is a beta version of the new Protege available. We should test it before the official release. What is our plan?
Ticket on another tracker
OBI ticket: https://github.com/obi-ontology/obi/issues/963
Ontology Developers' workshop
The meeting will focus on (Depending on who can attend):
- creating good logical definitions in both a general context and in the context of the GO-Reactome-Rhea alignment
- Filling in missing GO-Rhea xrefs
- moving from reactions to biochemical pathways
- implementation of a method to keep in synch with all three resources
- GO-CAM templates for pathways??
- potential implementation of Design Patterns and the revival of some type of TermGenie ability
- GH ticket work and mini-project planning
- Dealing with and migrating binding terms
- Dealing with cellular processes (https://github.com/geneontology/go-ontology/issues/12849)
- Taxon constaints- How do we handle them going forward?
- Attendees
- David H (Can make it)
- Kimberly (Can make it)
- Karen
- Harold (Can make it)
- Pascale (Can make it)
- Barbara (Maybe make it)
- Alan B (Rhea, can make any day)
- Anne (Rhea, prefers Monday or Tuesday)
- Chris
- Peter D?
- Ben?
- Jim?
- Paul T
Tickets
Failures due to use in other ontologies
Friday after much discussion, Rachael and I tried to finally obsolete lamina reticularis. When we created the pull request, the Travis check failed due to the use of the term in Uberon. Do we really want this behavior? It will make us dependent on other ontologies to proceed with our work.
https://travis-ci.org/geneontology/go-ontology/builds/425727676?utm_source=github_status&utm_medium=notification https://github.com/geneontology/go-ontology/pull/16345
GH project link
https://github.com/geneontology/go-ontology/projects/1
Minutes
- On call: Chris, David, Eric, Harold, Jim, Karen, Kimberly, Laurent-Philippe, Pascale, Tanya
Montreal GOC Meeting
- Add name to attendees list if you're going
Editors Discussion Topics
Order of Operations for External Ontologies
- We create xrefs to resources like Rhea, EC - when we need new xrefs, we should leave the ticket open to remember to add the xref, but shouldn't wait to add terms to GO
- Manually assert placement of terms in the ontology, but if needed, request terms in ChEBI and then Rhea to complete the logical definition
RO Meeting
- Pre-meeting tomorrow for October meeting
- Jim will present in October on causal relations
New beta version of Protege 5.5
- GO editors should test the new version before release
- Is there a sandbox that can be used for testing purposes?
- Should be able to install the new version and perform normal operations to make sure the new version works for all editing steps
- Specifically, check new obsoletion plugin - is it there?
Ticket on another tracker
- OBI ticket: https://github.com/obi-ontology/obi/issues/963
- One solution would be for OBI to just have their 'binding' be a child of BFO:process rather than GO Biological Process
- Jim will respond to the ticket
Ontology editors meeting - Geneva, December 2018
- Still need to work out the specifics of the agenda and who should be there for how long
Pathways
- Meet with Rhea group to coordinate xrefs between the two groups
- Need to coordinate Noctua development with ontology development, specifically wrt templates for pathways
Design Patterns and Term Genie
- We need to discuss good design pattern practice and whether we can/should resurrect Term Genie
Binding Terms
- Create proposal for Montreal meeting
- Sanity check in Geneva
Cellular Processes
- Re-factor - could there also be a proposal for Montreal?
- Sanity check in Geneva
Editors discussion github tickets
Obsolete GO terms in external ontologies
- Obsolete terms in external ontologies, e.g. UBERON, cause GO travis checks to fail
- For this case, it resulted in an inability to obsolete the GO term
- This check requires GO and the external group(s) to coordinate
- Does the term mean the same thing in GO and UBERON?
- The anatomists and the cell biologists may disagree on this term
- Chris removed any axioms in UBERON that referred to the GO term; we'll need to regenerate the UBERON import then
- Besides UBERON, the Cell Ontology and GO overlap
- But not ChEBI, for example, that does not use GO in its axioms
- We should also consider how GO is being used in phenotype ontologies that may use GO terms in more specific ways, e.g. 'heart development'
Remove all obsolete terms from slim subsets
- When a term is obsoleted, editors should also remove it from all subsets
- Like annotations, editors could alert people that a term is used in subsets so people can update their slims, if needed
Merge identical complexes present in different CCs
- Yes, if they are the same complex
- No, if they have different functions