starting from Neuron 7.0, impedance affects simulation!

Anything that doesn't fit elsewhere.
Post Reply
patoorio
Posts: 81
Joined: Wed Jan 30, 2008 12:46 pm

starting from Neuron 7.0, impedance affects simulation!

Post by patoorio »

Hi,

I want to monitor the impedance of the model neuron while simulation is running, let's say every 50 ms. So I made a script that advances 50 ms, pauses the simulations and recalculates an impedance object that is inserted. However, starting from Neuron 7.0 I found out that I can't do this any longer, as something (I don't know what yet) changes when the impedance is computed and the rest of the simulation is completely different!
I have a model of slow oscillation bursting. It is only one section, one segment, and has an impedance object inserted. If I do:

Code: Select all

** run for 2000 ms
imp.compute(0)
** run for 2000 ms
The last 2000 ms of the simulation will have a completely different pattern of oscillation. This is regardless of the presence of spiking (i.e. fast sodium and potassium) conductances. Looking at the oscillations and how they changed it is as if the leak (passive) conductance was decreased. However, it doesn't happen when the parameters are changed such that oscillations don't occur (i.e., the model is stable).
But the most important aspect of this problem is that it didn't happen with Neuron 6.2. What has changed since with the impedance class? I can still keep working with neuron 6.2 for a while, but I think this is something that should be addressed in some way... or not?

Best Regards

EDIT: You can reproduce it in a section with HH mechanism, and a IClamp that makes it fire continuously.
ted
Site Admin
Posts: 6299
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: starting from Neuron 7.0, impedance affects simulation!

Post by ted »

Sounds like a bug. There were several bug fixes between 7.0 and 7.1, the current version. Have you tried with 7.1?
patoorio
Posts: 81
Joined: Wed Jan 30, 2008 12:46 pm

Re: starting from Neuron 7.0, impedance affects simulation!

Post by patoorio »

Yes, it happens both with 7.0 and 7.1
hines
Site Admin
Posts: 1687
Joined: Wed May 18, 2005 3:32 pm

Re: starting from Neuron 7.0, impedance affects simulation!

Post by hines »

I believe
http://www.neuron.yale.edu/hg/neuron/nr ... ac38c7223f
has the changes required to fix that bug. Please let me know if it is insufficient in
your case. I also believe your current version would work if the dstate'/dstate contributions
are included in the calculation when Impedance.compute(freq, 1) is used.
patoorio
Posts: 81
Joined: Wed Jan 30, 2008 12:46 pm

Re: starting from Neuron 7.0, impedance affects simulation!

Post by patoorio »

Yes!! .compute(freq,1) did it.
I don't know how to try the code you posted, do I have to compile the whole program? Or there is a way to compile just those two files?

Thanks a lot!
hines
Site Admin
Posts: 1687
Joined: Wed May 18, 2005 3:32 pm

Re: starting from Neuron 7.0, impedance affects simulation!

Post by hines »

If you use those two files to replace the ones in nrn/src/nrniv then in that directory it should be sufficient to
make install
Post Reply