Ontology meeting 2013-07-18

From GO Wiki
Jump to navigation Jump to search

Attendees: Becky, Paola, Heiko, David, Tanya, Chris, Jane

Minutes: Tanya

ChEBI paper

Final version is submitted.

Homeostasis of number of cells vs. cellular homeostasis

Prompted by these links in assert job - Build 35:

ADD GO:0016057 'regulation of membrane potential in photoreceptor cell' GO:0070050 'neuron homeostasis'

ADD GO:0019228 'regulation of action potential in neuron' GO:0070050 'neuron homeostasis'

A proposal to fix the mix-up causing these is outlined here: https://www.ebi.ac.uk/panda/jira/browse/GO-92

I've removed the two added links for now, but they're bound to come back until we fix this :-)

TO DO: David will tease out which of our terms belong in which category and complete GO-92

Taxis, tropism and kinesis terms - and their link to 'axon guidance'

Following discussion at the SF call last week, Paola opened a SF ticket with suggestions to work on these terms. Please be brave, examine it and add comments.

https://sourceforge.net/p/geneontology/ontology-requests/10261/

TO-DO: Jane will review SF item. Others pls. also look

Issue with disjoint violation

Stemming from this SF request and discussion at last week's call: https://sourceforge.net/p/geneontology/ontology-requests/10239

David would like to discuss this - The violation is: 'formation by host of specialized structure for nutrient acquisition from symbiont' (equivalent to nothing)

TO-DO: David will delete developmental parentage from the parents of this term and leave the other 
parents and check if that solves the problem. DONE: problem solved.

Protein complexes

How to make sub-categories under protein complex. What is our preferred axis of classification - by function? Protein family? Process?

PREFERENCE: function axis first, then process, protein family less preferable

New relationships to connect process and components

We need some new relationships for relating components to processes in xps...we'd like has_participant and has_input for cases where a virus disrupts the function of a host complex but these would obviously be useful elsewhere too.

TO-DO:HJD is using 'capable_of' for his relationships between complex terms and molecular functions. We could
use 'capable of part of' for linking complexes and processes. HJD to review his file and make sure that use of 'capable of'
and 'capable of part of' are used correctly. 
TO-DO:Becky will make a SF item to illustrate the problem. We can all read and comment.

Protege issues?

David reported extreme slowness in loading some files. Has anyone else experienced this? In David's case, the slowness is much reduced provided that nothing else is running on his machine. Chris advices loading files locally rather than over the net.

Paola - loaded go.owl from purl, took 5 minutes, can live with that. Had to re-install P4.3 for some reason. Hermit took >10 minutes to run, then stalled. Downloaded Elk, version 0.3.2 January 2013. Runs fast but gives error:

UnsupportedOperationException: OWL API reasoner method is not implemented: getBottomObjectPropertyNode().

'positive regulation of execution phase of apoptosis' shows no inferences to 'positive regulation of apoptotic process', but can't be sure given reasoner fails/errors. Advice? Any further plugins I'd need?

TO-DO: Address while at JGI today. SOLVED:  David had an extra catalog.xml file that was causing problems.
Need document that describes the 'known' error messages that appear when running Protege that we can just ignore. 

Propagating via positive and negative regulation

We'd like to look at how links propagate via positively_regulates and negatively_regulates.

TO-DO:Examine propagation over 'part of' ? Will look at this later today. Can also use Ruth's SF ticket for an
example to test.

Bumped from last meeting: Molecule-level processes vs. population processes

How do we want to express these in logical definitions?

Stems from https://sourceforge.net/p/geneontology/ontology-requests/10142/

Think about other cases where there are molecule-level processes vs. population of molecule processes and
how we might tease these apart. This issue can be closed. 

Continues from last meeting: Protein binding/ protein domain binding xps

Background from last week: Chris did some work on logical xps for protein binding and protein domain binding. Need to talk about this some more.


Updates on ongoing work

Two directories in /scratch with initial work. Look at go-rhea-xref-cardinality.rpt, there are 75 one-to-many (about
half relate to NAD(P)H). Demo by Chris in Protege with file /scratch/rhea/mf-merged.owl.  May not need a Rhea link for
every reaction.