Padraig Gleeson wrote:

See comments in http://www.opensourcebrain.org/projects/muscle_model/repository/revisions/master/entry/NeuroML2/ca_boyle.nml.

The parameters in the paper suggest this is an activation variable, and since it's very slow (tau=151ms) it will mean the channel will hardly ever open...

In the title I think you mean "f" is activation not inactivation? I agree with you on this point. There is something very perplexing here, The activation rate would be too slow to support the kind of regular spiking we see in the data. Could this be a typo in the paper? I haven't looked at the c++ code.