E_pas

[Meena]

Hi,

Just a real basic question.

Assuming I am trying to give a section passive parameters, ie Rm and Cm.
Rm is basically in the units ohms.cm2

So using "insert pas"
my g_pas = 1/Rm
(Rm is obtained from literature)

I know e_pas is the equilibrium potential of the passive current.
so it would be ideal to set e_pas to v_init ??

Sorry...I know this is rather basic....but I just want to check anyway..

Thanks,
Meena

[ted]

[quote="Meena"]would be ideal to set e_pas to v_init ??[quote]
It would be OK but not necessarily ideal in any sense. In fact, it's the tail wagging the dog.
Think about the roles of e_pas and v_init. The former is supposed to reflect the biophysical
properties of a particular ionic current. The latter is the initial condition for a simulation. If
your aim is to ensure that the simulation starts at the resting potential of a section that has
only the pas mechanism, why wouldn't e_pas be the governing parameter? That being the
case, you would set v_init to the value of e_pas.

[Meena]

Hi again,

You mentioned :

" If your aim is to ensure that the simulation starts at the resting potential of a section that has only the pas mechanism, why wouldn't e_pas be the governing parameter?
That being the case, you would set v_init to the value of e_pas."

So if only the pas mechanism is present, then the resting potential is determined by that mechanism. And thus v_init should be set to e_pas?

So assuming now I have a more complex structure :
Situation 1:

A soma with several ion channels, exchangers and pumps, and 3 passive dendrites with Rm and Cm values. Lets say the resting potential of this neuron is -65mV?

Would I be correct in setting the dend.e_pas to be V_rest?

Situation 2:
A soma with several ion channels, exchangers and pumps, and 3 passive dendrites with Rm and Cm values. Lets say I am clamping the voltage of this neuron at -100mV.

Would the e_pas here be -100mV ?


Thanks for all your time,
Meena

[ted]

So if only the pas mechanism is present, then the resting potential is determined by that mechanism. And thus v_init should be set to e_pas?

Yes, if you want the initial condition to be the resting potential of the section.

In answer to your two new questions:

First, the overriding principle to keep in mind:
Your conceptual model of the cell should dictate the toplogy, geometry, and biophysical
properties of the computational model that you implement. Otherwise, the computational
would not be a fair test of your conceptual model.

When biophysical properties are nonuniform in space (inhomogeneous, as some prefer
to say), it is unlikely that the resulting model will have a uniform resting potential. You
must then decide how to deal with this. Furthermore, models that involve ion
accumulation mechanisms typically require special attention to ensure that initialization
achieves the modeler's intent.

NEURON's default initialization is equivalent to voltage clamping all membrane to v_init
for a long time, then turning off the clamp at t = 0. If membrane current mechanisms are
nonuniform, unless you are very lucky or have taken special care in designing your
model, membrane potential throughout the model will almost certainly start to change
at t = 0. If ion accumulation mechanisms are present, concentrations are also quite likely
to start to change. Usually this is undesirable; people generally want the simulation to
begin from a stationary point, i.e. a situation analogous to a wet-lab experiment in which
the cell is allowed to reach a steady state before stimuli are applied. One way to deal with
this is to just let the simulation run for however long it takes for v and concentrations to
settle, before you start to apply stimuli. The two drawbacks to such an approach are:
1. the "incubation time" can be very long
2. is a completely unsatisfactory method if your conceptual model requires v and/or
concentrations to be uniform, or even to be nonuniform but with values that YOU
specify, not values that are mere accidental consequences of the parameters you
assigned.

If you have a copy of The NEURON Book, please see chapter 8 which presents very
effective strategies for initializing models with ion accumulation mechanisms and/or
nonuniform biophysical properties. If not, read this preprint of that chapter:
http://www.neuron.yale.edu/ftp/ted/book ... xedref.pdf