SEVClamp

thats_karlo · Post by **thats_karlo** » Fri Jul 21, 2006 7:55 pm

Hi!

To try SEVClamp ( downloaded from nrn/src/nrnoc/svclmp.mod ), i wrote the following code.

load_file("nrngui.hoc")
create soma
access soma

soma{
 nseg=1 L=38.42 diam=26 Ra=173 cm= 0.88 
 
insert pas    g_pas=3.79e-5  e_pas=-76.5

 insert hh2  // In base of Traub& Miles for hipocampal pyramidal cell

gnabar_hh2=0.1  gkbar_hh2=0.05 vtraub_hh2=-52
ena=50 ek=-100

} 

objref stim

soma stim= new SEVClamp(0.5)
stim.dur[0]=200
stim.amp[0]=-75
stim.dur[1]=400
stim.amp[1]=-55
stim.dur[2]=200
stim.amp[2]=-75
stim.rs=1


tstop=800
v_init=-75
celsius=34

when i run this program in oc> prompt everything is fine!
in the voltage graph, i see three level of voltage in soma.
But, when i change

oc>stim.amp[1]=-35
oc>run()

in the vlotage graph, i see a oscillation of voltage during second step!!

that's not what i except! I except for every value for stim.amp[1], (-75,-65,-55,-35,-25,..), to see a step function with constant value for every duration of SEVClamp, Am i right? or there is something that i don't know?

1-Could you help me to find out, what's gonig wrong, or where is my mistake, or my misunderstanding for this simulation?

2- If insted of SEVClamp, simply i use VCalmp, i don't see any oscillation! and it seems every thing is fine.

I really confused about SEVClamp and VClamp, i know that SEVClamp is single electrod and VClamp is two electrods. and "If your goal is to study the properties of a cell, use SEClamp" (https://www.neuron.yale.edu/phpBB2/viewtopic.php?t=505)

I guess you feel i really mixed! await for any advice

Regards,
Karlo

Post by **ted** » Sun Jul 23, 2006 10:39 am

I really confused about SEVClamp and VClamp, i know that SEVClamp is single electrod and VClamp is two electrods

There is no SEVClamp in NEURON. The only way there is an SEVClamp is if you make
a new mod file whose NEURON block contains the statement
POINT_PROCESS SEVClamp
and then compile that mod file.

nrn-x.x/src/nrnoc/svclmp.mod is the NMODL source code for NEURON's built-in SEClamp. If you compile that, and then start NEURON, you will get this error message
The user defined name, SEClamp, already exists

So I will assume that you copied svclmp.mod to a new file, then edited the new file and changed
POINT_PROCESS SEClamp
to
POINT_PROCESS SEVClamp

in the vlotage graph, i see a oscillation of voltage during second step!!

To tell you why, I need to be able to reproduce the phenomenon. To reproduce the
phenomenon, I will need more information than this:

insert hh2 // In base of Traub& Miles for hipocampal pyramidal cell

Exactly where is the mod file that defines hh2? ModelDB doesn't contain an entry that
corresponds to one of the papers by Traub & Miles.

"If your goal is to study the properties of a cell, use SEClamp" (https://www.neuron.yale.edu/phpBB2/viewtopic.php?t=505)

True.

thats_karlo · Post by **thats_karlo** » Sun Jul 23, 2006 11:56 am

Dear Ted,

Thank you so much for your attention!

For hodgkin-huxle model and active property of neuron i used hh2.mod from following link :
http://www.neuron.yale.edu/cgi-bin/view ... cx/HH2.mod

Yes, you are right. i changed name of the POINT_PROCESS to SEVClamp.

As i mentioned when i use VClamp, every thing is fine!

However just, in oc> prompt type oc>stim.apm[1]=-35 you will see oscillation in voltage graph.

await for you reply!

Yours,
Karlo

Post by **ted** » Sun Jul 23, 2006 1:35 pm

in the vlotage graph, i see a oscillation of voltage during second step!!

There is no oscillation. There is only a slight overshoot because of inadequate voltage
control. I will assume that this overshoot is what you are referring to.

The explanation is that when you use VClamp, the effective series resistance is given by
rstim/gain. Default gain for VClamp is 1e5, and rstim is 1 megohm, so effective series
resistance is 10 ohms. Compare that with the default rs of SEClamp (1 megohm) and you
will see why the latter allows some overshoot. To get similar voltage control with SEClamp,
just reduce its rs by a factor of 1e5. With most model cells I have seen, it is only necessary
to reduce rs to 0.001 to get quite adequate control.

A comment: Voltage control is a serious concern in experimental work. Patch clamp series
resistance reported by experimentalists is generally 10 megohms or greater. There are few
reports of experiments that managed to achieve smaller series resistances.

Post by **ted** » Sun Jul 23, 2006 1:43 pm

A few comments about the HH2.mod that you mentioned: it has some "loose ends"
that you may want to fix if you are thinking of using it in your own work.

In ASSIGNED block add units to declarations of tau_m, tau_h, and tau_n,
e.g.
tau_m (ms)

Also, to complete the conversion of that mod file from its "pre-cvode" form--

In NEURON block comment out RANGE declaration of m_exp, h_exp, n_exp
: RANGE m_exp, h_exp, n_exp

In PARAMETER block comment out declaration of dt
: dt (ms)

In PROCEDURE evaluate_fct comment out the lines that calculate m_exp, h_exp, n_exp
e.g.
: m_exp = 1 - Exp(-dt/tau_m)

thats_karlo · Post by **thats_karlo** » Sun Jul 23, 2006 2:21 pm

Thanks Ted.

I send to you an email(to this address ted.carnevale@yale.edu) with a *.jpeg attachment that show my voltage graph. it's really oscillation!

(how can i insert an image in the reply post. should it hase an internet address?)

await to heare from you again.

Regards,
karlo

Post by **ted** » Sun Jul 23, 2006 2:42 pm

thats_karlo wrote:it's really oscillation!

So it is. Your hoc code said
insert hh2
but the mod file declared
SUFFIX hh2ad
When I changed the insert statement to
insert hh2ad
I neglected to change the suffixes in the statements that specify the conductance densities.

Whatever.

Same cause, same fix.

thats_karlo · Post by **thats_karlo** » Sun Jul 23, 2006 3:13 pm

oc>run () ---- in votage graph every thing is fine
oc>stim.amp[1]=-35 --------> oscillation! (as you have seen)

i really don't find my mistake!! help me please!!!

Post by **ted** » Sun Jul 23, 2006 5:55 pm

You wanted to know why the membrane potential oscillated when you used an SEClamp
and not when you used a VClamp. This is the answer

ted wrote:The explanation is that when you use VClamp, the effective series resistance is given by
rstim/gain. Default gain for VClamp is 1e5, and rstim is 1 megohm, so effective series
resistance is 10 ohms. Compare that with the default rs of SEClamp (1 megohm)

In other words, a VClamp with rstim == 1 megohm and gain = 1e5 is closer to being an ideal
voltage source than an SEClamp with rs == 1 megohm is. To do a "fair" comparision
between the two, you would need to reduce the SEClamp's rs to 1 megohm / 1e5
= 10 ohms. If you do that, you will find that the SEClamp controls membrane potential
quite well.