This is the readme for the model associated with the paper:

Uebachs et al. Efficacy loss of the anticonvulsant carbamazepine in
mice lacking sodium channel beta subunits via paradoxical effects on
persistent sodium currents, in revision.

These model files were supplied by Marie-Therese Horstmann.

Usage:
------

Check that you have NEURON installed (available from
http://www.neuron.yale.edu). Autolaunch from ModelDB **or** download
and extract the archive and compile the mod files to run under ...

linux/unix
----------
by typing
nrnivmodl
in the top level directory.  Then type
nrngui mosinit.hoc

mswin
-------

Run mknrndll, cd to the expanded directory and press make nrnmech.dll
button.  Double click on the mosinit.hoc file.

MAC OS X
--------

Drag and drop the expanded folder onto mknrndll icon in the NEURON
application folder. Drag and drop the mosinit.hoc file onto the nrngui
icon to start the simulation.

Once the simulation is running:
-------------------------------

Click the "Run simulation/graph then write file" button to recreate
figure 9A from the paper:

screenshot

The simulation takes 7 minutes on a linux 2.8GHz pentium 4. You can
check whether it's running or not by looking at the "t" parameter
which shows the current "model time". You can stop the simulation by
the "stop" button. There is a 500ms pre-iteration period before the
current stimulation. The model needs this time to establish a steady
state. The membrane voltage trace from soma is written in a file
called "outputdata.dat" in the directory, where the model is
located. This is a plain ascii file. One data point corresponds to
0.25ms.  You can also, of course, change the model parameters varied
in the publication above.