recording a variable from a module

[Prendergast]

I apologize if this is an ill-informed question, but how can one record a variable used only in an individual module of a simulation (in my particular case, glutamate concentration)? When I try to use vector.record(&variable), NEURON produces an "undefined variable" error. Thanks in advance for your help.

[ted]

Please provide the NEURON block of the mechanisms whose glutamate
concentration you are trying to record, and the exact hoc statements that
create the Vector and tell hoc to record glutamate concentration to that
vector.

[Prendergast]

My glutamate module is unfinished, pending experimental results, but I've been using another module to make sure that I'll be able to obtain data when it is complete. The test module is:

Code: Select all

TITLE passive membrane channel

NEURON {
	SUFFIX pasglu
	NONSPECIFIC_CURRENT i
	USEION Glu READ Glui VALENCE 1
	RANGE g, e, Gluinf, multi, gglu, gmulti
	GLOBAL Glui, gglu
}

UNITS {
	(molar)	= (1/liter)
	(mV) = (millivolt)
	(mA) = (milliamp)
	(mM)	= (millimolar)
	(S) = (siemens)
}

PARAMETER {
	g = .001	(S/cm2)	<0,1e9>
	e = -70	(mV)
	Gluinf = 2.4e-4
	multi = 0
	gmulti = 0
	gglu = 0
}
STATE {
	Glui	(mM)
}
ASSIGNED {v (mV)  i (mA/cm2)}

BREAKPOINT {
	evaluate_fct(v,Glui)
	i = (g-gglu)*(v - e)
}

PROCEDURE evaluate_fct(v (mV), Glui (mM)) {
	:gglu=multi*(1-Glui/Gluinf)*g
	gglu = multi * (1 / ( 1 + exp(-(gmulti*(Glui-Gluinf))/0.005) )-1/2) * g
}

and the hoc code is:

Code: Select all

objref gluconcntr
gluconcntr = new Vector()		//to record glutamate concentrations
gluconcntr.record(&gglu)

[ted]

When the mechanism is complete, how do you envision using it? That is, do
you want it to produce a single "synaptic activation" in the course of a
simulation run, or do you want it to be activatable more than once per run?
And do you need to be able to emulate the attachment of multiple glutamatergic
synapses onto the postsynaptic cell so that they are electrotonically close to
each other?

[Prendergast]

The simulation is of spindle activity in the thalamus over the course of about ninety seconds, so it incorporates many activations. It's almost complete, and multiple synapses are already incorporated. The problem is that my PI is interested in the inhibitory effects of glutamate on spindle activity, and wishes to be able to keep track of glutamate levels as the simulation progresses.
At the moment, glutamate concentration is computed in a separate module, but I'm having trouble recording those data. As I mentioned, I have limited experience in the NEURON environment, so I may me overlooking a simple fix. Thanks for your help!

[ted]

The answer to your question depends strongly on how your model is
implemented, at a level of detail that has not yet entered this discussion.
If you can send me
ted dot carnevale at yale dot edu
a toy model that contains just the essential hoc and
mod files to analyze the implementation of synaptic transmission, and how
it operates in a test simulation, I will try to advise you how to proceed.