Pinsky & Rinzel version of Traub model

Discussions of particular models.

Moderator: tom_morse

Pinsky & Rinzel version of Traub model

Postby pdennevis » Tue Mar 28, 2017 6:29 am

I am diving into the model of Pinsky and Rinzel and I face some questions I hope you are able to help me with.

First of all, I have seen different approaches in modeling neurons. P&R (and after them Schweighofer in 1999) omit specific compartment dimensions and calculate current densities using the compartment's surface area ratio, whereas other models, for instance in the NEURON environment use explicit compartment lengths and radii.
--> Why would they in the first approach choose to divide by compartment area? What do they gain? What do they lose?


Second, I would want to make a model that does implement the cell's dimensions from the P&R model. In order to do so, I have to choose the correct parameters.
ted wrote:You have to figure out what surface areas your sections must have, and what conductance and capacitance densities they should have, and you have to choose the correct section diameter and length and Ra so that the series resistance between your compartments will be the same as P&R used. It's a pain but it's the way things are.

--> What is the best approach to find these values? Should I just test a range of values and look which set of parameter values looks the most like the P&R model or is there something else you suggest?

Looking forward to your response!
pdennevis
 
Posts: 3
Joined: Fri Dec 09, 2016 11:37 am

Re: two compartment Traub model, share membrane voltage

Postby ted » Tue Mar 28, 2017 12:24 pm

P&R (and after them Schweighofer in 1999) omit specific compartment dimensions and calculate current densities using the compartment's surface area ratio
because they're dealing with one- or two-compartment models, and as mathematicians/theoreticians they value simplicity and clarity.
other models, for instance in the NEURON environment use explicit compartment lengths and radii
because NEURON was designed to meet the needs of experimentalists, who cannot dispel the complexities of real neurons (such as their complex branched architectures) by a wave of the hand and a theorem.
I would want to make a model that does implement the cell's dimensions from the P&R model. In order to do so, I have to choose the correct parameters.
. . .
What is the best approach to find these values?
The details depend entirely on how the original model was implemented, and the requirements of the tool that you are using to implement your own replica of that model. Since length and diameter are essential to NEURON, use whatever lengths and diameters are specified in the original implementation (taking into account the fact that NEURON uses microns for units of length). If the original implementation did not specify any lengths or diameters, it's up to you to choose appropriate values for your reimplementation. After that, you have to figure out how to map each of the remaining parameters into a form that is suitable for use with NEURON. This a tedious process that requires you to know the units of each parameter in the original implementation, what they should be in the reimplementation, and what to do when they don't match up. In principle it's no worse than converting between metric and US units for length, mass etc.. If your starting point is source code that does not document the units assumed for parameters and variables, refer to the original publication that describes the model. Also, don't forget that some modelers and theoreticians like time to be in seconds, while others prefer milliseconds.
ted
Site Admin
 
Posts: 5040
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine


Return to Specific models

Who is online

Users browsing this forum: No registered users and 1 guest