Add new reduced model types

[rgerkin]

@JustasB and I plan to add new reduced model types (like izhikevich2007Cell) after a survey of the literature to figure out which reduced models would be useful.

[pgleeson]

Sounds good! @rgerkin @JustasB, I've added you to a team under NeuroML, so you should be able to self assign issues like this...

[rgerkin]

First step: list all of the reduced neuron model types that have been expressed in NeuroML in one place or another.

[rgerkin]

@pgleeson
Two links here are broken (one for the definition and one for running it). I'm not sure if one of these is meant to link to one of the files here or not, but once I see your fix for one of these models, I should be able to fix others on my own.

Also, what is the difference between the "WCell" and the "SCell" version.

[rgerkin]

I will eventually be making versions of each of these models with dimensions, which will be more compatible with testing.

[rgerkin]

List of reduced models defined in NeuroML2 core types:

• Integrate and Fire
  • Defined without refractory period as iafCell
  • Defined with refractory period as iafRefCell
  • Also see iafTauCell and iafTauRefCell for versions with time constant instead of capacitance parameters.
• Izhikevich
  • Defined without units as izhikevichCell
  • Define with units as izhikevich2007Cell
• Adaptive Exponential Integrate-and-Fire
  • Defined with units as adExIaFCell
• Fitzhugh-Nagumo
  • Defined without units as fitzHughNagumoCell
  • There is also supposed to be a fitzHughNagumo1969Cell but the link is broken here
• Pinsky-Rinzel CA3
  • Not actually "reduced" but probably a generalization of simple conductance-based models.
  • Defined with units as pinksyRinzelCA3Cell

List of models implemented in OSB projects:

• Morris-Lecar
• Hindmar-Rose
• Hodgkin-Huxley

List of unimplemented (or undiscovered) models of interest:

• Quadratic IF
• See also this list of reduced models in PyNN.

[pgleeson]

@rgerkin I've updated the Morris Lecar wiki and now it just grabs the content of the README at https://github.com/OpenSourceBrain/MorrisLecarModel. I've fixed the links. The "WCell" and the "SCell" versions are from Wikipedia and Scholarpedia, and are pretty much the same. It would be great if someone can double check them, make a call about which is more appropriate and then it can be put into the core types.
A dimensional version would be good too, but the dimensionless version might be closer to what's been more widely used.

[pgleeson]

See also: https://github.com/OpenSourceBrain/AllenInstituteNeuroML/tree/master/CellTypesDatabase/models/GLIF