Ontology meeting 2015-09-10

From GO Wiki
Jump to navigation Jump to search


Minutes: David OS

DC meeting action item review

AI (DH): Merge all 'X production' to 'regulation of production'.  
We keep this as regulation of a process, rather than changing to regulation of a quality: 
The thing measured (level) is a quality, but that is the results of regulation of some (grouping) process.
The only drawback for this is that it doesn't work for axiomatisation to PRO without using nested
class expressions - not compatible with OBO.
AI (DOS): send around email for protein complex working group:
 Darren (PRO), Sandra  & Birgit (IntAct), Paola + David OS + Helen + Melanie + Chris (GO), Peter (Reactome).

MF refactoring

 Action item (Melanie to bring to manager call):
 GO PIs + Chris M. to determine whether resources can be provided for required improvements 
 to general MF axiomatisation prior to refactoring. This mainly consists of pulling in mappings to ChEBI 
 from Rhea - possibly from other resources via EC.  Could be simple subclassing axioms if full logical definitions prove 
 problematic.  Without this work we will not be able to automate classification under new classes requested by Paul T.
 Action item: DOS to make trello ticket
 Suzi and Paul T to provide examples from LEGO of why refactoring is needed.
 Other action items on hold until the above items are sorted. 

User experience (UX)

Action item:training video for how to create a slim - build from existing training talks? Claire mentioned that EBI is making videos. Action item:create a tool that would use a biology concept (angiogenesis) to select their slim terms (of the various slim flavors, this is the flavor made by gathering more granular terms related to and around the seeding term)

 * This is slim in a broader sense than just 'terms with subset tag'.  
 * A couple of ways to automate based on concepts of interest to Pharma:
 ** Use axiomatisation of GO to create appropriate grouping classes as for the Roche CV mapping project
 ** Create groupings based on co-annotation (same GP, same paper).  P2G already suggests co-annotated terms.
 AI: Melanie to get details of this P2G funcationality of P2G and then email editors (or wider?) with details.
 AI: DH to look into how this might be done in Amigo

TermGenie Review backlog

Any additional input needed from the group for any of the terms in the queue?

 Secretion term(s): suggest reg of production instead
 Production term(s): : suggest reg of production instead
 Response to - where assaying results of poisining rather than endogenous, evolved response:  
   Allow terms if it makes sense that there is an endogenous, evolved response, but add comment warning of usage.
   Email requester advising against usage if paper clearly does just assay toxicity.

  'sequence-specific double-stranded DNA binding'  Is this needed?
  A. Yes:  Add & also add single stranded.  Most of existing heirarchy will live under new term
  Comb through hierarchy for anything the might be SS.  Try to find some examples from lit:
  Harold has idea for example -  send to DOS.
  Review of SS DNA binding proteins here: http://www.ncbi.nlm.nih.gov/pubmed/22976174

Why is ligand-gated ion channel activity not under receptor activity ?

These are generally referred to as receptors by biologists. Was this a deliberate choice, if so, what is the reason behind it?

See https://github.com/geneontology/go-ontology/issues/12029

  No reason for this.  Just an omission.  Move ligand-gated ion channel under receptor activity. 
 Also consider light activated ion channels.  Should these be receptors too?

Jira tickets

Time for our periodic review of Jira tickets, especially unassigned ones. Please go through JIRA issues in advance of meeting. We will aim to devote about 20' of the call to discuss Jira tickets.