dllbuild error: Method cnexp can't be used...

[andy_t_roo]

the error:
Method cnexp can't be used with Block scheme at line 56 in file caBuffer.mod

i'm trying to implement calcium Buffering (i know there is an example tutorial with shells.)

the relevent parts of the model

BREAKPOINT {
SOLVE scheme METHOD cnexp
}

DERIVATIVE states {
cai' = (minf-cai/cb)/mtau + ica/(FARADAY*svRatio)
}

does anyone know what can cause that error?

thanks,
Andrew

ps.
does anyone know of any NEURON models that simulate ionic concentration change?

i'll put the full model file at andrewhill.no-ip.org/caBuffer.mod
(it's not there at the moment - i'll do it in an hr or so)

[ted]

If this is true

the relevent parts of the model

BREAKPOINT {
SOLVE scheme METHOD cnexp
}

DERIVATIVE states {
cai' = (minf-cai/cb)/mtau + ica/(FARADAY*svRatio)
}

the fix is pretty clear: change scheme to states and the error msg should go away.

does anyone know of any NEURON models that simulate ionic concentration change?

Do you want the concentration to change _before_, or _during_, a simulation?

[andy_t_roo]

ted,
thanks for pointing out the obvious mistake - how on earth could i have missed that.

I'm interested in watching how the the ionic concentration of the different ions changes during the simulation - you have ions moving in and out of the cell and i would be interested in seeing the long term effects on the ion ballence of adding a new ion channel, or changing the properties of the existing ones.

thanks again,
Andrew

[ted]

thanks for pointing out the obvious mistake - how on earth could i have missed that.

Everybody does that kind of thing. In fact I just did it yesterday on two separate
occasions myself.

I'm interested in watching how the the ionic concentration of the different ions changes during the simulation - you have ions moving in and out of the cell and i would be interested in seeing the long term effects on the ion ballence of adding a new ion channel, or changing the properties of the existing ones.

You need an ion accumulation mechanism for each ionic species in question. Such a
mechanism has a USEION x statement that READs ix and WRITEs xi and/or xo,
where x is the name of the ionic species (na, k, ca, cl etc., although for cl and others
you need to also specify VALENCE). For examples, see "Expanding NEURON's
Repertoire of Mechanisms with NMODL" (get it from the Documentation page at
www.neuron.yale.edu). An updated version of that will appear in chapter 9 of The
NEURON Book. If you're using MSWin, c:\nrnxx\examples\nrniv\nmodl has several
ion accumulation mechanisms. UNIX/Linux/OS X users will find these in
the source code (I always build from sources, and on my Linux box they're in
/usr/local/src/nrn-5.8/share/examples/nrniv/nmodl )