Signaling Meeting Minutes July 2009

From GO Wiki
Revision as of 10:30, 30 October 2009 by Jdeegan (talk | contribs)

Jump to: navigation, search

13th July 2009

Participants: Alex Diehl, Ruth Lovering, Peter D'Eustachio, Harold Drabkin, Sandra Orchard, Susan Tweedie.
Chair: Jennifer Deegan
Minutes: Jennifer Deegan
File: go/scratch/signaling3.obo

Agenda items:

  • Demo of new top term structure.
  • Which areas of signaling in GO most urgently need attention?
  • Does anybody have time to meet for one hour a week to work on edits?

Questions raised by Nicolas Le Novere before the meeting:

Some of the issues I wanted to bring on were:

1) "neuronal signal transduction" is_a "signal transduction via ionic flux".

This is not correct, and translates the fact that one term describes a phenomenon at the molecular level, while the other describes a phenomenon at the cellular level. First it is not clear what "neuronal signal transduction" means. Is-it a neuron that transduces the signal coming from other neurons to its targets (including signal propagation and integration)? Or is-it the transduction of the signal sent by a neuron to the next? For the latter, there are plenty of different types: Volume transmission, wired transmission, that itself can be through an electrical synapse, a chemical synapse etc. A chemical synapse can in turn use ligand-gated ionic channels, GPCR, growth factors receptors etc.

2) "signal transduction via ionic flux"

I believe this term is overloaded and does not represent what you meant. A ionic flux is just another type of diffusible mediator. Think about calcium for instance. This is presumably not what was intended here. What we need to have is something like:

"signal transduction via the change of an environment parameter"

It could have the following children:

"signal transduction via the change of membrane polarisation"
"signal transduction via the change of osmotic pressure"
"signal transduction via the change of pH"

3) I still think that beside the diffusible mediator and the environment parameter, we need two other types:

"signal transduction via conformational change", as it is the case for the upper part of GPCR cascade for instance. But also much of the signal transduction involved in cellular movement, deformation, etc.

Something like "signal transduction via clustering". This is how all the growth factor receptors function. Once activated (generally through phosphorylation, itself brought by clustering of the receptors), they cause the aggregation of downstream targets, increasing their local concentration and causing signal propagation. Examples are EGFR, NGFR, TGFbR etc.

Do-we need something like

"signal transduction via covalent modification"? The phosphorylations are generally only a way to get a conformational change or to create a handle for clustering.

Meeting Minutes

  • Renamed "signaling" to "signaling process"
  • Need place to list parts of signling and types of signal.
  • Added new term "signal termination"
  • Considered "signal transmission" v. "signal transduction" v. "signaling process"
  • Does "signal initiation" = "signal detection"? - No. Signals are not always detected.
  • Is a nerve impulse a signal? Yes, any transfer of information is a signal. But maybe that is too high level a way of thinking about it? Nerve impulse is very different from e.g. ion or hormone signaling. If we try to think at this high level at the beginning then we may get bogged down. Perhaps best to start with one small easy area?
  • Perhaps we could make a structure like "reproduction" and "reproductive process". That way we can put the parts of signaling under one and the types under the other.
  • We should separate the initiators from the transducers.
  • With the "signal initiation" process can we annotate the hormone or just the receptor? - Yes we should also annotate the hormone. The membrane-bound components including the ligand and receptor can all be annotated to this. We should then break down by class in child terms. e.g. G-protein bound.
  • We should define the start the end of each process.
Example: Insulin

Assume that insulin already exists so we don't need to worry about biosynthesis and release?

Immunologists are very interested in where the signal comes from which contradicts the first assumption. 

There are 3 or 4 possible consequences depending on who the signal 
talks to e.g. ras protein. Insulin receptor is tyrosine kinase and 
G-proteins are distinct from that. We should look for existing relevant terms and slot them into the hierarchy. 
  • We need to think about time.
    • Keep "signal initiation" and "signal transduction" separate. Initiation happens at the membrane. {Not always, e.g., nuclear hormone receptors, where initiation can occur in the cytosol or the nucleoplasm. While progression along most signaling cascades is indeed associated with change of location in the cell - from membrane to cytosol to nucleus - it's not clear that this association is true in all cases or that it is straightforward enough to be part of a good general definition.)
    • There may be several different end points from a single different signal initiator.
    • Many different signals may set off G-proteins. The effect depends on which G-protein is set off.
  • Ruth finds a picture of insulin signaling and reads out the parts...[1]
    • It is all protein kinases, not G-proteins.
    • We need to take into account cell type and cell state.

Signal initiation summary

  • The "signal initation" branch of the ontology may be quite shallow.
  • We will annotate the signal, the receptor, and the complex that is formed when they bind.
  • What about things that affect the strength of the response? Should we have a regulation term for that?
  • There is an isoform example where the working end of the cellular receptor is missing and so though it uses up all of the ligand, no signal is passed on. This is an example of a regulator.
  • "Signal initiation" is a very short part of the signaling process. It is important that we don't throw everything under initiation. This term should just include the part of the process where the ligand binds the receptor. Everything else goes under "signal transduction". For example passing or amplifying a signal goes under "signal transduction".
  • We can't currently put insulin (or any other ligand, e.g. fibroblast growth factor) under "signaling". This is why we need the initiation term, as this is the step in the overall process that involves the molecular function of the ligand.
  • Another step that belongs under "signal initiation" is the step where the resulting complex reorganises or activates. There is a conformation change, or an autophosphorylation so that the complex can now interact productively with downstream elements. All of this can also go under "signal initiation". The inert receptor may be in the membrane, the cytosol (e.g. some nuclear hormone receptors), or the nucleoplasm (e.g. some other nuclear hormone receptors).
  • Start and finish of "signal initiation" have now been defined. Start is ligand binding to inert receptor. End is the end of the conformation change to produce the activated receptor.
  • Some initiation steps may have a transduction component but this is not likely to be a problem. Nuclear receptors will also not present a problem.

Next meeting

We need to think about the boundary between "signal initiation" and the next step. Can we get good examples?

The next meeting will be on Monday 20th July at: 8a.m. PDT, 9a.m. MDT, 10a.m. CDT, 11a.m. EDT, 4p.m. BST, 5p.m. CEST. We will again use Clackpoint and webex as these technologies worked well.

20th July 2009

Participants: Ruth Lovering, Peter D'Eustachio, Harold Drabkin, Sandra Orchard, Susan Tweedie.
Chair: Jennifer Deegan
Minutes: Jennifer Deegan
File: go/scratch/signaling3.obo


We would like to identify the boundary between "signal initiation" and the next step. Can we get good examples?


signal transduction v. signal transmission

Does signal transduction imply change?

No, we decided that signal transduction is a phrase that is commonly used as an exact synonym for signal transmission, even though in technically correct language they have different meanings. We have decided not to make separate terms for signal transduction and signal transmission as this would be too confusing in the light of common usage. Instead we will make a term "signal transduction" and give it the exact synonym "signal transmission".

We will also keep regulation of "signal transduction" terms separate from "signal termination".

signaling v. signaling process

  • Terms defined:

"signal termination"
"signaling process" (def not complete)

Action Item: Complete the definition of "signaling process".

  • Order of sub-processes

The sub-processes include:

"signal initiation"
"signal transduction"
"signal amplification"
"signal propagation"
"signal termination"

These processes do not necessarily happen in this order. All we can say is that the end of termination must be last and the start of initiation must be first.

  • The end of signal initiation is the point when a receptor becomes active.
  • The end of signal termination is the point when the receptor becomes inactive again.
  • signal termination can end before the end of the whole signaling process.

Action item: We need to add an end pooint to "signaling". We do not want to accidentally include translation and transcription just because of not having been clear about the end of signaling.

regulation terms

  • Agonists: We need to distinguish between positive regulation during amplification, and positive regulation feeding back from the end of the process to amplification step.
  • Added some regulation terms under "signaling process" and put the pre-existing "regulation of signal tranduction" terms in place under these.
  • Where should we annotate the creation of the signaling molecule? It should be annotated to the biosynthesis term. For example the creation of insulin should be annotated to "insulin biosynthesis".

Next meeting

The next meeting will be on Monday 3rd August at:
8a.m. PDT, 9a.m. MDT, 10a.m. CDT, 11a.m. EDT, 4p.m. BST, 5p.m. CEST.
We will again use Clackpoint and webex.

23rd July 2009

Sandra Orchard has suggested that I should put all terms with name "X signaling pathway" under "signaling" so that we can go through them at the next meeting and check which belong there. She says that many of these terms just have regulation child terms and so will be fine as they are, but some have actual process child terms and so will need some further work. I am moving the terms today in preparation for Monday's meeting. Having moved the terms I also removed all resulting redundant relationships, and I move the new grouping term "old signaling terms to be rehomed" and put it as a part_of child under signaling. This enabled the grouping term to be ignored in the redundancy calculation.

28th July 2009

  • A question from Barry Smith:

If 'signal transducer activity' does not mean 'function', then what is the difference between it and BP:signal transduction?

  • Answer from David Hill:

Well, if we were to follow what Tanya and I proposed in Eugene, we would turn 'signal transducer activity' into 'molecular function involved in signal transduction'.

Q/ Do we agree?

3rd August 2009

Participants: Ruth Lovering, Alex Diehl, Sandra Orchard, Jennifer Deegan

Top terms

We are still considering the top terms, how the are defined and how they relate to one another.

"Signaling process" is a grouping term under which we plan to corral all of the parts of signaling.

"Signaling" is intended as a term under which we can group all _complete_signaling_pathway_ terms. Signaling pathway terms each define the whole of one signaling system, for example, "insulin signaling pathway" or "notch signaling pathway".

We discussed at length the problems of terminology.

We have a lot of rather general terms and we are not sure how we want to use them all:

  • Signaling process
  • Signaling pathway
  • signal transduction
  • signaling

The pathway term is thought to include everything in one given signaling progression from arrival of the ligand to the end of the effect on the cell. An example of a pathway is the insulin signaling pathway.

We considered whether we should delete the word 'pathway' from the end of the signaling pathway terms but decided not to since it is in common usage across the literature.


We are also considering how these relate to the whole:

  • celllular signaling
  • multicellular organismal signaling
  • Intracellular signaling
  • intercellular signaling
  • endocrine signaling
  • autocrine signaling


We discussed signaling adaptation and signaling regulation and how they related to one another. The adaptation term seems to be about a pathway regulating itself in a long term way, whereas the general regulation term may be about other things regulating a signaling pathway.

Is the ligand part of the pathway?

We came round again to the question of whether a signaling pathway include the biosynthesis of the ligand. It was suggested that we should consider broadening the definition of the the top "signaling" term so that it includes the biosynthesis of the ligand, and that it should also be general enough to cover all other types of signaling. However, as this was contrary to decisions made in previous meetings it was thought that this should be discussed when more participants were present. We decided that we would need to have a neuronal signaling expert present for this discussion and Jennifer will ask Nicolas le Novere if he can come next week to discuss this.

We considered whether the ligands are made when the signal needs to be sent, or whether ligands are stockpiled in advance. It seems as thought interleukin is an example of signal that is stockpiled.

In insulin signaling it was suggested that glucose detection is not part of insulin signaling.

This last topic was one we had discussed and agreed on before and there was some concern that we were wasting time by rehashing things we'd already agreed. For future discussions it would probably be best if we only discuss big issues like this when everyone is present so we do not come back to points we have already discussed and agreed at previous meetings. I will try to get people to let me know earlier if they will be able to be present for meetings, so we only go ahead when we have quorum.

Next week

We are still working on top terms. We will try to have enough people present to define the top signaling term completely. Jennifer has invited Nicolas le Novere.

2nd September 2009

Participants: Ruth Lovering, Alex Diehl, Peter D'Eustachio, Jennifer Deegan, Sandra Orchard, Susan Tweedie.

Signal biosynthesis

Is the biosynthesis of a signal part of the signaling pathway?

  • Pathway almost always starts after ligand binding.
  • Signal transmission is movement of a signal from one place to another, either across the body or across the cell.
  • Release of the signal happens at different times, so we can't bundle it together with the signaling pathway.
  • A receptor gene product can be annotated to a response term. For example, the insulin receptor protein could be annotated to response to glucose.
  • Peter: Finding the boundaries in pathways is notoriously hard. We find this repeatedly in Reactome. The boundaries must be kept very narrow or it gets too complicated. For example, digestion process is affected by not eating, but you would not make ingestion a part of digestion.
  • Human insulin signaling: Someone with type I diabetes would not be labelled as not carrying out the insulin pathway.

Signal transmission

Note: we had previously decided that signal transduction and signal transmission should be one term, as this is the state of common usage in the literature. However at this point, this decision is reversed.

Alex: Maybe signal transmission is transit of a signal across the body or cell. Signal transduction is transit across the cell. Therefore signal transduction is an is_a child of signal transmission.

Action: We created and defined signal transmission and rearranged the graph, including the initiation and termination terms.

Structure under signal transmission

Will signal transmission have any children?

  • Signal transmission needs a child term 'signal release' e.g. for cytokines.
  • Signal transmission groups all of the parts, like release and transduction; all processes adjacent in time.
  • This will also let us support different views of the boundaries of pathways.

Structure and defs agreed today are as follows:


id: GO:0007165
name: signal transduction
namespace: biological_process
def: "The process whereby an activated receptor conveys information 
down the signaling pathway, resulting in a change in the function or
state of the cell. In many cases the end of this process is the 
initiation of transcription." [GOC:go_curators]

id: ID:0000027
name: signal transmission
namespace: biological_process
def: "The process whereby a signal is released and/or conveyed from one
location to another location within an organism. The process terminates
at the end of the sub-process of signal transduction, when the function
of the cell has been altered as a consequence of receipt of the signal."

id: ID:0000028
name: signal release
namespace: biological_process
def: "The process whereby a signal is secreted or discharged into the 
extracellular medium from a cellular    
source." [GOC:mtg_signaling]

id: ID:0000000
name: initiation of signal transduction
namespace: biological_process
def: "The process whereby a signal causes activation of a receptor,
for example, via a conformation change. Gene products participating 
in this process include both ligands and receptors." [GOC:mtg_sig]

These defs remain as:

id: ID:0000018
name: signaling
namespace: biological_process
def: "The entirety of the process whereby information is sent from one 
location to another within a living organism. The synthesis of the signal 
is not, in general, considered to be part of this process. For example 
synthesis of insulin is not considered to be part of the insulin signaling 
pathway, and the insulin is assumed to have been made before the signaling 
process starts. In a signaling process that goes to conclusion, the final 
step will be a change in the function of a cell." [GOC:mtg_signaling]

id: ID:0000012
name: signaling process
namespace: biological_process
def: "Any biological process involved in the generation, transmission, 
reception, or interpretation of a signal. A signal is an entity used 
to transmit or convey information." [GOC:mtg_signaling]

28th October 2009

To be written...