integration of HH channel gating states

Anything that doesn't fit elsewhere.
Post Reply
brocke
Posts: 11
Joined: Fri Aug 16, 2013 9:21 am

integration of HH channel gating states

Post by brocke »

Hi,

do I understand correct, that HH-like gating states are *always* integrated using analytic method independently which method I specify in the SOLVE statement in the mod file? If it's wrong, how to manipulate the integration method?

thank you in advance,
e.brocke
ted
Site Admin
Posts: 5784
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: integration of HH channel gating states

Post by ted »

brocke wrote:do I understand correct, that HH-like gating states are *always* integrated using analytic method independently which method I specify in the SOLVE statement in the mod file?
Your understanding is incorrect. How did you arrive at it?
If it's wrong, how to manipulate the integration method?
First let's address the important question that you managed to not ask: which integration method to specify in the SOLVE statement? Unlike so many other things in contemporary life, it's not a matter of preference--there really is a correct answer. See
Integration methods for SOLVE statements
in the Forum's Hot tips area.

Options for "manipulating the integration method" are limited. The available choices are fixed time step and adaptive integration. Fixed time step integration offers a choice between the default fully implicit Euler method and two Crank-Nicholson variants (read about secondorder in the Programmer's Reference). If you resort to adaptive integration, NEURON automatically selects something from a collection of methods based on CVODES or IDA; depending on the particular model, you may be able to specify whether each model cell in it advances synchronously with all the others (global variable time step) or at its own pace (local variable time step). For more information see chapter 8 of the NEURON Book and the brief discussion under What is NEURON? at http://www.neuron.yale.edu/neuron/about.
brocke
Posts: 11
Joined: Fri Aug 16, 2013 9:21 am

Re: integration of HH channel gating states

Post by brocke »

Thank you for a quick reply.
brocke wrote:Your understanding is incorrect. How did you arrive at it?
I tried the default SOLVE statement and I tried SOLVE statement with cnexp to solve HH like channel gates. While debugging in both cases the channel gate variables were calculated by the integration of analytic solution generated in the corresponding hh.c file (static int states () function).
ted
Site Admin
Posts: 5784
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: integration of HH channel gating states

Post by ted »

If the dynamics of gating states are specified by one or more ordinary differential equations, then a
SOLVE ... METHOD cnexp
statement causes automatic generation of code that works correctly for all of NEURON's integrators. In particular, with fixed time step integration, the values of the gating states are calculated from the analytical solution to the ODEs, but with adaptive integration the ODEs themselves are solved by numerical integration.
Post Reply