gap junction example doesn't work...

[blop]

Hi,

I wanted to implement a gap junction in my model and failed. So I went back to the NEURON book, created an empty test directory, copied the gap.mod file :

Code: Select all

NEURON {

	POINT_PROCESS gap
	POINTER vgap
	NONSPECIFIC_CURRENT i
	RANGE r, i
}
PARAMETER {
	r = 2e9 (megohm)
}
ASSIGNED {
	v (millivolt)
	vgap (millivolt)
	i (nanoamp)
}
BREAKPOINT {
	i = (v - vgap)/r
}

as well as the book's example hoc file

Code: Select all

create a,b
access a
forall {nseg=10 L=1000 diam=10 insert hh}
objref g
g = new gap()
g.r =3 

and I obtain the following error:

Additional mechanisms from files
gap.mod
/Applications/NEURON-6.0/nrn/i386/bin/nrniv.app/Contents/MacOS/nrniv: point process not located in a section
in test.hoc near line 6
g.r =3
^

Now, if I do locate the synapse (that's what I did in my main model) by using

Code: Select all

create a,b
access a
forall {nseg=10 L=1000 diam=10 insert hh}
objref g
g = new gap(0.5)
g.r =3 

The script runs well... until I use finitialize or fadvance

finitialize(-50)
/Applications/NEURON-6.0/nrn/i386/bin/nrniv.app/Contents/MacOS/nrniv: Bus error See $NEURONHOME/lib/help/oc.help
near line 1
finitialize(-50)
^

NB: I'm under Mac OS 10.4 and use NEURON 6.0

[ted]

Except that what's in the book does work.

Your version of gap.mod isn't a copy of what's in the book; however, that's
not the cause of the problems you encountered.

as well as the book's example hoc file

Take another look at page 272 and compare it with what you posted here.
(Come to think of it, make sure that what you posted here is what you tried
to use on your Mac.)
What is posted here is
--lacking one of the objrefs
--doesn't assign a location to the objref that it does have
--doesn't link the POINTER variable vgap to the "presynaptic" membrane potential

The moral of this story: steal working code (we like to call this "programming
by example," which sounds so much more genteel, no?), but don't break it
in the process.

[blop]

What is posted here is
--lacking one of the objrefs
--doesn't assign a location to the objref that it does have
--doesn't link the POINTER variable vgap to the "presynaptic" membrane potential

The first point wasn't the problem I guess since I tried that too

Code: Select all

create a,b
access a
forall {nseg=10 L=1000 diam=10 insert hh}
objref g[2]
for i=0,1 {
	g[i] = new gap()
	g[i].r =3 
	}

and it gives the same error message,
which means that the end of the file

Code: Select all

a g[0].loc(0.9999)
b g[1].loc(0.0001)
setpointer g[0].vgap, b.v(0.0001)
setpointer g[1].vgap, a.v(0.9999)

is needed. And I hardly understand this since I though that the commands (except loops, procedure, etc) could be typed in the oc> directly, i.e. linearly. I guess that objects belong to the "etc" category of my parenthesis...

The moral of this story: steal working code (we like to call this "programming
by example," which sounds so much more genteel, no?), but don't break it
in the process.

Well, when I'm stealing code, I try to understand what I steal. So I play with it to see if it follows my understanding. Call this "torture the examples" if you want ;-)

[blop]

PS: I tried to torture this example a bit further with something like this:

Code: Select all

begintemplate myneuron
   public soma, gapsyn
   create soma
   objref gapsyn
  
   proc init() {
      create soma
      objref gapsyn
      gapsyn=new gap(0.5)
      }
endtemplate myneuron

And

Code: Select all

xopen("myneuron.tem")			
objref N1,N2
N1 = new myneuron()
N2 = new myneuron()

setpointer N1.gapsyn.vgap, N2.soma.v(0.5)
setpointer N2.gapsyn.vgap, N1.soma.v(0.5)

access N1.soma

finitialize(-50)

and I have to admit that, at my great surprise, it works...

[EtOH_ataxia]

OK, I need help.
I am trying to do exactly what blop did. Here are the files I made, to me they look the same as blop's working version.
Except they don't work. I get an error saying

nrniv: gap is not a template
in myneuron.tem near line 9
gapsyn=new gap(0.5)
^
xopen("myneuron.tem" )

I don't really understand, I thought gap should not be a template!

The only difference I can tell is the operating system, I am using NEURON 6.1 on Windows XP.

Code: Select all

begintemplate myneuron
   public soma, gapsyn
   create soma
   objref gapsyn
 
   proc init() {
      create soma
      objref gapsyn
      gapsyn=new gap(0.5)
      }
endtemplate myneuron

Code: Select all

xopen("myneuron.tem")
objref N1,N2
N1 = new myneuron()
N2 = new myneuron()

setpointer N1.gapsyn.vgap, N2.soma.v(0.5)
setpointer N2.gapsyn.vgap, N1.soma.v(0.5)

access N1.soma

finitialize(-50)

Code: Select all

NEURON {
POINT_PROCESS gap
POINTER vgap
NONSPECIFIC_CURRENT i
RANGE r, i

}
PARAMETER { r = 1e10 (megohm) }
ASSIGNED {
v (millivolt)
vgap (millivolt)
i (nanoamp)
}
BREAKPOINT { 
i = (v - vgap)/r 
}

Sorry, but I can't find the bug...can you help me?

[EtOH_ataxia]

I got a version that works
I tried the previous files also with Linux (Zenwalk -Slackware) and they still did not work. Same error.
the version that works for me is this (same .hoc, only changed r to 1 to see a large effect)

Code: Select all

objref gapsyn_1, gapsyn_2

create cell_1

cell_1{
insert hh
}

create cell_2

cell_1 gapsyn_1=new gap(0.5)
cell_2 gapsyn_2=new gap(0.5)

setpointer gapsyn_1.vgap, cell_2.v(0.5)
setpointer gapsyn_2.vgap, cell_1.v(0.5)

I made cell_1 generating an action potential (by lowering gkbar_hh to 0.003) so the v waveform was obvious
I still do not understand why it is not working with the template version, though.
Paola

[tmwilliams]

I'm having similar problems. I'm using NEURON version 7.0 and MSWindows. My code is as follows:

My gap.hoc file is:

Code: Select all

create a,b

access a

forall {nseg=10 L=1000 diam=10 insert hh}

objref g[2]

for i=0,1 {

	g[i] = new Gap()

	g[i].r = 3
}


a g[0].loc(0.9999)	

b g[1].loc(0.0001)  

setpointer g[0].vgap, b.v(0.0001)

setpointer g[1].vgap, a.v(0.9999)

My gap.mod file is:

Code: Select all

NEURON {
	POINT_PROCESS Gap
	POINTER vgap
	RANGE r, i	
	NONSPECIFIC_CURRENT i
}

PARAMETER { r = 1e10 (megohm) }

ASSIGNED {
	v (millivolt)
	vgap (millivolt)
	i (nanoamp)
}

BREAKPOINT { i = (v - vgap)/r }

As far as I can tell, my code is identical to that in The NEURON Book. I compiled the .mod file using mknrndll.

When I try to run gap.hoc, I get the following error:

nrniv: point process not located in a section
in C:/Documents and Settings/tmw/Desktop/gap.hoc near line 14
}
^

Can someone tell me what the problem is?

[ted]

Can someone tell me what the problem is?

Must be what the interpreter said--

Code: Select all

point process not located in a section

a is the default section, but it appears that, for versions of NEURON more recent than what I used when writing this example for Chapter 10, an actual location must be specified when a point process is created. Change the statement

Code: Select all

   g[i] = new Gap()

to

Code: Select all

   g[i] = new Gap(0.5)

and the example in the book works. Score another erratum for the NEURON Book.

[tmwilliams]

Thanks for the explanation, I almost went crazy yesterday trying to figure this out!

I'm sorry, but I'm really new to this NEURON software, so I have a couple more questions.

I'm assuming that the statements:

Code: Select all

a g[0].loc(0.9999)   

b g[1].loc(0.0001) 

setpointer g[0].vgap, b.v(0.0001)

setpointer g[1].vgap, a.v(0.9999)

change the pointer location from 0.5 to 0.9999 for cell a and 0.0001 for cell b. Is that correct?

And, can you tell me what the following statement means?

Code: Select all

g[i].r = 3

[ted]

change the pointer location from 0.5 to 0.9999 for cell a and 0.0001 for cell b

Almost. Variables that are continuous functions of position in the original biological system--which are called "range variables" in NEURON (e.g. membrane potential)--are computed at locations that are equally spaced along the interior of each section. Each section has its own discretization parameter nseg that specifies how many nodes are used to represent the section.

Code: Select all

create dend // nseg == 1 by default, node is at 0.5
print dend.diam(0.5) // diameter at middle of dend
dend.nseg = 3 // nodes are at 1/6, 1/2, and 5/6
print dend.diam(x) // if 0<x<1, prints diam at nearest node to x

a g[0].loc(0.9999) places g at the internal node that is closest to the 1 end, assuming that a.nseg is not very large.

Since you are new to NEURON, it would be advisable to read chapters 4 and 5 of The NEURON Book ASAP (your original question suggests you do have the book; if I have jumped to the wrong conclusion, then at least read the "Key papers about NEURON" listed on NEURON's Documentation page http://www.neuron.yale.edu/neuron/docs).

can you tell me what the following statement means?

Code: Select all

g[i].r = 3

Assign 3 to the parameter r that belongs to the object referenced by g. Here's how to fish: have a go at chapter 13 of the NEURON book, or at least read http://www.neuron.yale.edu/neuron/stati ... n/obj.html

[tmwilliams]

I think I didn't word my question clearly, I understood that the statement: g.r = 3 was assigning the value 3 to the parameter r that belongs to g, I just couldn't figure out what r was. Later I realized that it is (obviously) the resistance, I must have been having a senior moment. Thanks for your help!

[ted]

A recurring puzzle in any Forum or mailing list is how to interpret questions--deciding whether a question requires a literal answer, what pre-existing knowledge may be assumed and what must be stated explicitly, what level of detail to get into etc..