Demo

thats_karlo · Post by **thats_karlo** » Tue Jul 25, 2006 2:29 pm

Hi!

To compare my simulation with other result i download a demo(ofcourse a NEURON code; ModelDB ). The arthure claims that he has obtained his results from this code.
I run this demo, in oc> prompt i type oc>forall psection()

The information about property of model (section and their properties) are exactly like my model. I have used his *.mod files also. According to demo, I insert a current Clamp with the same parametrs.

But result is defferent!!!! (form and ferequecy of spikes are different)

I guess there are (is) some hiden parameters that dose not appear after oc>forall psection()

Does it true? if yes, how can we access them?

Post by **ted** » Tue Jul 25, 2006 6:46 pm

Try the Model View tool. On this page
http://www.neuron.yale.edu/neuron/docs
you will find a link to a presentation about it.

thats_karlo · Post by **thats_karlo** » Tue Jul 25, 2006 9:53 pm

Hi!

I have tryed Model View for this demo, but it causes some error, and all windows are closed!

Post by **ted** » Tue Jul 25, 2006 11:26 pm

If you want further help with this, please specify exactly what "demo" you were
playing with. A URL that points directly at it would be most useful.

thats_karlo · Post by **thats_karlo** » Thu Jul 27, 2006 4:36 pm

Dear Ted,

From Tools/Distributed Mechanism/Manager, i find the difference between my model and Demo file!

In contrast to their articles, they have used other numbers for some parameters!!!! to obtain good response.

However, thank you for your attention,

Yours,

Karlo

thats_karlo · Post by **thats_karlo** » Mon Aug 07, 2006 1:28 pm

Dear Ted,

Again i get some difficultis with demo file. i send my code and address of demo file to your email address "ted.carnevale@yale.edu".

I hope, you find time to look at my code and that demo file,

await for your reply,

Yours,
Karlo

Post by **ted** » Tue Aug 08, 2006 5:02 pm

First, i check the input resistance of my code and Demo file,(for a current clamp I=0.075nA)

oc>objref zz
oc>zz = new Impedance()
oc>soma zz.loc(0.5)
oc>zz.compute(0)
oc>rn = zz.input(0.5)

- The result for demo file is 104.75157 and for my code
is 104.75115.(I don't know where this error is coming from,
maybe numerical precision?)

The difference is pretty small, but if it's genuine, maybe it is caused by one or more of
the following:

1. Different model specification. Are you sure that the anatomical and biophysical properties of your model are identical to the model in
> http://senselab.med.yale.edu/senselab/m ... ?model=279
? Don't trust the values that are displayed in the Parameter panels, which tend to round
off at 5 significant figures. Actually print the values from the hoc command line.

Also make sure that the values of nseg are identical (but I think that a difference in nseg
would have produced a much bigger difference in results).

2. Different "instrumentation" or experimental conditions. Are you sure that you are
using the same temperature (celsius) and that all current and/or voltage clamps have the
same amplitudes and durations?

3. Different initialization, or inconsistent initialization. Is the initial membrane potential the
same? If ion accumulation mechamisms are present, are the concentrations identical
after initialization? If you click on the RunControl's Init button several times, and calculate
impedance each time, do you get the same impedance each time?

4. Different integration time step or run time. Is dt the same? Is t identical at the end of
the simulation?

thats_karlo · Post by **thats_karlo** » Tue Aug 15, 2006 9:52 am

Dear Ted,

Are you sure that the anatomical and biophysical properties of your model are identical to the model in.Actually print the values from the hoc command line.Also make sure that the values of nseg are identical (but I think that a difference in nseg would have produced a much bigger difference in results).

Yes, I have checked this parameters.

Different "instrumentation" or experimental conditions.

They are equall. Even i test input resistance without any current or voltage clamp.

Different integration time step or run time. Is dt the same? Is t identical at the end of the simulation

It has checked!

Different initialization, or inconsistent initialization.

what i underestant from initialization is V-init should be the same. right?
when i want to applay some IClamp or VCalmp, i run the protocol after 100ms to be sure, i have a steady state. i don't know is it a good way or not!

If ion accumulation mechamisms are present, are the concentrations identical after initialization

.

Yes, in my model we have a ion accumulation mechanism. from
Tools>Distributed Mechanism>Viwers> Mechanism(Globals)
i check concentration of ions.

They don't change during any initialization and even during run the code.
(i guess you mean other test that i don't know what is that?)
so the answer for

If you click on the RunControl's Init button several times, and calculate
impedance each time, do you get the same impedance each time

is negative.

Could you give me some hints about initialization,(i have read chapter of neuron bookabout this subject!).

I have other question, that i will sent to your email address.

Yours,
Karlo