Extending Single-Compartment to Multi-Compartment model

NMODL and the Channel Builder.
Post Reply
Kvogt
Posts: 11
Joined: Mon Mar 25, 2019 5:34 pm

Extending Single-Compartment to Multi-Compartment model

Post by Kvogt » Wed Jun 26, 2019 12:15 pm

We are trying to extend a single compartment model to a multi-compartment model to validate stimulation behavior. The single compartment model is based on the HH model with adjusted channel parameters. In order to extend this model to the multi-compartment model, we used the following setup:
In the soma, we inserted the Ra, cm and adjusted HH mechanism
In the axon, we inserted the Ra, cm and adjusted HH mechanism
In the apical dendrites, we inserted Ra, cm and pas mechanism
In the basilar dendrites, we inserted Ra, cm and pas mechanism
We also want to inject current for stimulation so we added a Point Process.

Is that a proper way to extend a single compartment model to a multi-compartment model? And if not can you please suggest a right one or references on how to extend the single compartment modle to multi compartment model?

Thank you in advance.

ted
Site Admin
Posts: 5632
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: Extending Single-Compartment to Multi-Compartment model

Post by ted » Thu Jun 27, 2019 2:10 pm

Your model is your model. What you include in it, and what you omit, is up to you to decide. Such decisions should be based on the purpose that motivated implementation of the model in the first place.

When a modeling project reaches a point at which the original model has been studied and conclusions inferred, the attention often shifts from whatever the original question was to a new question that is some variant of "how generalizable are the conclusions that were inferred from the original model?" This is where the modeler can thoughtfully relax one or more of the simplifying assumptions that went into the design of the original model.

But note that this can introduce problems. In going from a single compartment model to a model that has significant spatial extent (e.g. two or more branches, some of which may have different ion channels) it is easy to end up with something that does not have a uniform resting potential, so NEURON's default initialization does not put the model into a steady resting state. Your very briefly described model seems prone to that problem. For example, the hh mechanism by itself has a resting potential of -65 mV, but the default resting potential for a section that contains pas is -70 mV. Attach a section with pas to a section that has hh, and the result is a model in which the local resting potential is neither -70 nor -65 mV but somewhere in between, depending on the values of g_pas, Ra, and the anatomical dimensions of the model. If you change e_pas to -65 mV in every section that has pas, this problem will not occur.

You don't say what "adjusted HH mechanism" means, so I have no basis for suggesting what will happen in the model that you propose.

Kvogt
Posts: 11
Joined: Mon Mar 25, 2019 5:34 pm

Re: Extending Single-Compartment to Multi-Compartment model

Post by Kvogt » Tue Jul 02, 2019 10:27 pm

Dear Ted,

Thank you for responding.

The modified hh mechanism has adjusted parameter infinity curves and time constant values. The above neuron setup was based on the hh model from the NEURON book. We are curious about the contributions of the various mechanisms. Namely, what do the Ra and cm mechanisms do, and do you know why the book might have these mechanisms in all of the compartments? What does the pas mechanism do, and why would it be added to only the dendritic compartments? We are trying to understand why the book would use the mechanisms in a setup like this to extend the hh mechanism to see if a similar approach would work for us.

ted
Site Admin
Posts: 5632
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: Extending Single-Compartment to Multi-Compartment model

Post by ted » Wed Jul 03, 2019 2:41 pm

We are curious about the contributions of the various mechanisms. Namely, what do the Ra and cm mechanisms do, and do you know why the book might have these mechanisms in all of the compartments?
Ra and cm are not mechanisms. They are attributes of sections. A section is used to represent the properties of an unbranched neurite.

If you have the NEURON Book, let me point you to chapter 5, which will answer your questions. If you don't have the book, or if you do have the book but want a quicker introduction, read
The NEURON Simulation Environment
https://neuron.yale.edu/neuron/static/p ... /nctoc.htm

Kvogt
Posts: 11
Joined: Mon Mar 25, 2019 5:34 pm

Re: Extending Single-Compartment to Multi-Compartment model

Post by Kvogt » Wed Jul 17, 2019 6:27 pm

Dear Ted,

thank you for your reply.

Post Reply