Modeling a Test Iclamp

The basics of how to develop, test, and use models.
Post Reply
charles1
Posts: 24
Joined: Thu Apr 26, 2012 8:41 am

Modeling a Test Iclamp

Post by charles1 »

Hi,

I have a morphologically reconstructed cell via a cellbuilder. I wish to create a single test or model pulse of given v_init, delay, amplitude, duration, and total time in Iclamp for this cell. I then wish to download the resultant V(t) vector, in order to model it as per the MRF2 tutorial. Except that here, the data is not experimental, but from NEURON.

Thank you.
ted
Site Admin
Posts: 5795
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: Modeling a Test Iclamp

Post by ted »

So what's your question?
charles1
Posts: 24
Joined: Thu Apr 26, 2012 8:41 am

Re: Modeling a Test Iclamp

Post by charles1 »

I apologize for the lack of clarity.

How is what I wish to achieve done?

I tried exporting the voltage trace from the voltage axis graph by using "pick vector"; the resolution is very low - not enough voltage points, and no time points. I tried changing dt in the run control menu; the trace is more refined, but "pick vector" yields the same result. So "pick vector" is not the right way. What is?
ted
Site Admin
Posts: 5795
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: Modeling a Test Iclamp

Post by ted »

charles1 wrote:I tried exporting the voltage trace from the voltage axis graph by using "pick vector"; the resolution is very low - not enough voltage points, and no time points.
Very strange. After I "pick" a vector (left click on a trace in a graph that is in "Pick vector" mode),
NEURON Main Menu / Vector / Save to File
writes a text file that starts like so

Code: Select all

label:node[5].v( 0.5 )
201
0       -70
0.025   -70
0.05    -70
 . . .
i.e. first line tells the name of the dependent variable, 2nd line how many xy pairs, and the remaining lines contain time and the dependent variable as the first and 2nd fields on each line.
I tried changing dt in the run control menu; the trace is more refined, but "pick vector" yields the same result.
dt doesn't specify how many points are plotted per ms. That is controlled by a standard run system parameter called steps_per_ms. At initialization, the standard run system first makes sure that dt is compatible with steps_per_ms, and if it isn't, forces dt to an appropriate value. If you want to change steps_per_ms, you could do so with a simple hoc assignment statement, or you could enter the desired value into the numeric field next to the RunControl's "Points plotted/ms." Example: by default, steps_per_ms is 40 and dt is 0.025 ms. Doubling steps_per_ms changes dt to 0.0125 ms. If I run a new simulation, pick a new vector and save it to a file, I find that the data are sampled at intervals of 0.0125 ms.

If you prefer to capture the time course of a variable directly to a Vector, use the Vector class's record() method (documented in the Programmer's Reference entries about the Vector class). Of course you will then have to capture t to a separate vector, and write your own code to ensure that dt and the recorded variable are written to one or more text files.
Post Reply