mod files and GUI

Using the graphical user interface to build and exercise models. Includes customizing the GUI by writing a little bit of hoc or Python
Post Reply
fskinner
Posts: 3
Joined: Fri Jun 10, 2005 12:36 pm
Location: Toronto, Canada
Contact:

mod files and GUI

Post by fskinner »

Hello,

I am in the process of converting to using the GUI, so 2 quick questions at this point:

1) It seems that it is better (i.e., computationally efficient etc.) to create channels via the GUI channel builder rather than the "old-fashioned" way of mod files. Is this an appropriate assumption? (as we proceed)

2) Otherwise, to use existing mod files, I can compile them fine, but then not sure how to insert them into my morphology via GUI (which is there via a loaded .hoc file or loaded in Cellbuilder).

Thanks for help!
fskinner
Posts: 3
Joined: Fri Jun 10, 2005 12:36 pm
Location: Toronto, Canada
Contact:

Re: mod files and GUI

Post by fskinner »

don't worry for my 2nd quick question - I was making it more complicated than necessary - just had to have the mod file in the directory.

however, my first query still stands - would you recommend using the GUI channelbuilder rather than continuing with mod files?
ted
Site Admin
Posts: 6289
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: mod files and GUI

Post by ted »

Good questions.
fskinner wrote:1) It seems that it is better (i.e., computationally efficient etc.) to create channels via the GUI channel builder rather than the "old-fashioned" way of mod files. Is this an appropriate assumption?
The computational speed advantage of Channel Builder-specified mechanisms over mechanisms specified with NMODL is small--probably not significant compared to other things that can be done to speed up program setup and execution when run time is really an issue. If you already have channels specified with NMODL code that is "tried and tested" I'm not sure it's worth the effort to reinvent them in Channel Builder form.

The Channel Builder does have the advantage that it avoids the many peculiarities of NMODL, while at the same time providing immediate graphical display of the voltage-dependence of rates or taus and inf values, and its built-in menu of (physically realizable) formulas prevents a lot of mistakes (at the cost of preventing the use of completely empirical formulas for voltage-dependence of rates). And it can be used to implement stochastic gating, much more easily than NMODL. But implementing temperature dependence is awkward (discussed elsewhere on the NEURON Forum). And for now we are still stuck with NMODL for accumulation or transport mechanisms, or adding new kinds of artificial spiking cells.
fskinner
Posts: 3
Joined: Fri Jun 10, 2005 12:36 pm
Location: Toronto, Canada
Contact:

Re: mod files and GUI

Post by fskinner »

thank you - will take this all into consideration as we move forth on various fronts.
Post Reply