January 10th 2011
As part of the monthly annotation calls, Rebecca presented three signaling WG proposals to the consortium:
- 1/ Removing the process/pathway split in signaling
-All signaling pathway terms to move back under 'signal transduction' -Signaling pathways start with ligand-receptor binding, and end with regulation of a downstream process. -Obsolete all signal transmission terms, as they are ambiguous and too broad to be useful -Create terms for endocrine and paracrine signaling (see minutes for discussison)
NB: This also includes the proposal to change 'x signaling pathway involved in y process' terms TO ' x signaling pathway involved in REGULATION OF y process' to be clearer that the signaling part ends with regulation of a downstream process, but not the process itself (see minutes for discussion).
Slides can be found here
REMOVAL OF THE PATHWAY/PROCESS SPLIT WAS COMMITTED TO THE ONTOLOGY ON FRIDAY 14TH JANUARY
- 2/ Removing subcellular location information from receptor terms.
Removal of the 'transmembrane' text string from receptor function terms. -Removing the 'TM' part from 'TM receptors' and suggesting that component information is captured with cellular component annotations instead.
Slides can be found here
- 3/ Removing the 'receptor signaling protein activity' terms.
-Create broader terms for 'signal transducing x kinase activity' -Move 'the enzyme receptors' under these terms (this removes the distinction between receptors and their downstream signaling molecules). -Merge the existing 'receptor signaling protein activity' terms into these new broader terms.
Slides can be found here
Present: Rama SGD; Jodie SGD; Kimberley; Karen WormBase; David MGI; Becky GO EBI; Emily GOA EBI; Yasmin GOA EBI; Pascale Geneva; Ruth BHF-UCL; Mike Princeton.
- Everyone agreed that the process/pathway split in signaling should be removed.
- Everyone agreed that the signal transmission terms can be obsoleted
- David H. asked that the physiologists should look at the 'endocrine signaling' term, to see if it can be well defined before adding it in (therefore need to hold off on the autocrine term too). Although he understands they are proposed to replace the 'signal transmission via vascular system' and 'signal transmission via diffusion' terms that will be obsoleted, in practise it's very difficult to define 'autocrine signaling'.
- David H. brought to everyones attention that receptor agonist under signal transduction covers ligand receptor function under specific pathways. Ligand receptor binding agonists should have relationship to receptor agonist and hence should become a child of -ive regulation of signal transduction.
- Becky - ligand receptor binding - individual parts of signal transduction process and pathways are all named. Each step will occur on parts of pathway or initiation of signal transduction or termination of signal transduction. A pathway is a named set of signal transduction. Wants to also create general new terms for endocrine signaling and paracrine signaling - we already have autocrine signaling. Wants to obsolete signal transmission.
- Rama asked about the proposed structure - is signal transduction in the same branch as responds to?
- Becky - all signaling terms are children of a cellular response to stimulus. |Signal transduction terms are broad at the moment and therefore useless eg: transduction via air transmission = transmission via diffusible molecule. Endocrine will be used to replace signaling transmission via vascularisation.
- David H - Stan needs to take this proposal to physiologists in Wisconsin.
- Backy - can Stan ask plant biologists if endocrine signaling can be used for plants?
- AI: receptor antagonist activity should be made a child of 'negative regulation of signal transduction'. (DONE)
- AI: check how the signaling terms fit with the 'response to' terms.
The second part of the proposal: changing 'x signaling pathway involved in y process' to 'x signaling pathway involved in REGULATION of y process'
- David H - this is in agreement with what was previously discussed i.e receptor signaling pathway is regulating thhe process it is happening in. That 'regulation of y process BY x signaling pathway may be better wording, and it would mean the term would get two is_a parents which is teh preferred.
- regulation of regulation terms are NOT allowed (there's one in the ontology from David H but we shouldn't be adding more).
- some processes are regulation terms themselves (eg l/r asymmetry determination) and it doesn't make sense for these to have 'regulation' children.
- For global processes (eg heart development), rather than having 'regulation of heart development by x signaling pathway', it would be better to link the pathway to regulation of a more specific process. E.g. 'regulation of transcription involved in heart develoment by x signaling pathway'. The regulation steps are often involved in the global process itself.
- AI: Becky will generate a list of all the terms that need to be changed, and talk to David about the changes she wants to make, to check it fits in with his and Tanya's work.
- AI: Becky will look back at presentations from Salt Lake city about 'regulation'.
PROPOSAL 2/: REMOVING TM NOMENCLATURE FROM RECEPTOR TERMS
- Becky - TM should be in cellular component ontology and not in function ontology; therefore would like to remove TM part of term name which will involve remaining TM function terms having TM removed from name. could add comment - consider annotating to integral part of membrane CC term. This will also apply to nuclear.
- David H pointed out that if the location is required for the function, then these could stay in as long as they have logical definitions. Becky emailed Chris M and he confirmed this. E.g. it's a receptor activity that spans a membrane. Or a receptor activity that transmits a signal across a membrane.
- Ruth L - pointed out that important information will be lost i.e nuclear receptor activity is a specific statement and is involved in nregulating transcription
- Karen - pointed out that there are problems nuclear receptor terms and there needs to be some changes.
- David H - also not all nuclear receptors reside in the nucleus
- Ruth L - no they reside outside the nucleus in cytoplasm and then make their way to the nucleus.
- Becky - maybe could have as synonym asnd remove nuclear from term name?
- Ruth L - need a statement to cover all nuclear receptors which bind to a hormone and then move to nucleus and regulate transcription.
- Rama - we do already have component specific terms e.g.mitochondrion
- Karen - is there a different way that the nuclear TM receptors behave?
- Ruth L - nuclear receptors are not in membrane, only in cytoplasm and then move through nuclear pores into nucleus
- David H. TM receptor activity will be receptor activity in the TM and therefore is covered by interontology link to generate cross product definitions.
- AI: Becky will email Chris M and improve the definitions of the TM receptors but will not remove the 'transmembrane' bit from the name for now. (DONE:Chris M confirmed that 'transmembrane can stay in the terms so long as they have improved logical definitions: Becky will update the definitions).
PROPOSAL 3/: CLEANING UP RECEPTOR SIGNALING PROTEINS
- Becky do receptor signaling protein terms belong in function?
Proposal to delete receptor signaling protein activity terms and replace with signal transducer terms. - can be captured in process of signaling transduction in downstream process.
- David H thinks the proposal works: anything that is annotated to 'signal-transducing protein tyrosine kinase activity' but NOT 'receptor activity' can be assumed to be a downstream molecule.
- Ruth L is concerned about removing the distinction between receptors and the downstream molecules. Becky isn't convinced this distinction belongs in function.
- Karen reminded Becky that the nuclear receptors need some work. These aren't always nuclear: they start off in the cytoplasm and enter the nucleus to regulate transcription.
- RuthL - everyone knows MAP kinases are intracellular - may need to have more subdivisions
- Becky - this is where interontology links will come into effect
- David H - will need to make it easier to disntinguish between the receptors and receptor activity in MF ontology
- Becky - anything that wasn't a receptor would be a downstream effect. Receptors would just renamed having got rid of TM in name.
- David thinks this may work
- Becky - will go down the transport route for receptor and function. Has-part never got resolved as to how to use in the function ontology?
- Karen - it has already been done for transcriprion factors so it is possible to use has_part in MF ontology.
- Rama - once the changes are in then we should go over 1 signaling pathway to see how it works
- Becky - this is what is planned for the 2nd day of the signaling workhop in Feb 2011. We will come up with guidelines on how to annotate to signaling terms - what proteins in the pathway can be annotated. Will go through in a future annotation conference call. In signal transduction the difference is that the ligand is part of signal transduction but is not a transducer.
- AI: Becky will look at these more carefully.