NMDA_S cannot be used with CVODE

[Feanor]

Hello to all,

I am trying to modify the file nmda2.mod file from http://senselab.med.yale.edu/ModelDB/Sh ... \nmda2.mod so that it writes the calcium current, ica, and that it responds to an event received by a NetStim. I get the error

Code: Select all

NMDA_S cannot be used with CVODE

I thought that a call to net_event() in the NET_RECEIVE block would be enough. Is it due to not having a call to/improper use of at_time() ?

The other modifications to the file are as follows:

1) nmda2.mod has a pointer pre which should be set to point to a variable depending on which release of transmitter is simulated. I have replaced that variable with an indicator variable saying whether an event has arrived. It is set in the NET_RECEIVE block:

Code: Select all

NET_RECEIVE(dummy (uS)) {
	indicator = 1
	net_event(t)
}

2) At line 179 of nmda2.mod I have replaced the if statement that checks whether pre is above the threshold for transmitter release with:

Code: Select all

if (indicator == 1) {		: spike occured?
			indicator = 0

3) The NONSPECIFIC_CURRENT statement has been replaced with

Code: Select all

USEION ca WRITE ica

Best regards

[ted]

First question for you: why is this worth the effort?

Doesn't ModelDB contain event-driven synaptic mechanisms that implement saturation and voltage-dependent Mg block? If not, it would be much easier to start with the event-driven "AMPA_S" mechanism described in chapter 10 of the NEURON book, add v-dependent Mg block, and change the time constants to whatever you want. And you'd have a mechanism that works with all of NEURON's integrators. And it would be able to handle multiple converging event streams--a very useful feature if your conceptual network architecture has more than one synapse of the same type located within the same compartment, because your computational implementation of the conceptual model can represent all of them with a single synaptic mechanism that is the target of multiple NetCons.

[Feanor]

The answer to the question - I didn't find any at the time, but yes, I didn't look thoroughly. Alternatively, yes, starting with the AMPA_S mechanism sounds like the way to go. Thank you!