explore response properties of individual HNN network cells

[cjayb]

On investigating the LFP-implementation (#150), I've come across the need to understand our cells a bit better. The work in #322 is hugely beneficial. The topic of "cell biophysics" is coming up here and there (#321), and there's passing mention to "updated calcium dynamics", which I assume is WIP elsewhere (perhaps a publication in preparation?).

I'm proposing to

• [ ] create examples of "virtual patching" of our cells (current injection at section A, intracellular recording of Vm at sections X, Y & Z)
• [ ] analyse the "dipole currents" (slight misnomer IMHO) inside cells due to synaptic inputs
• [ ] analyse axial (transmembrane) currents following AMPA and NMDA synaptic input to specific sections (e.g. 'apical_tuft')
• [ ] calculate extracellular potentials in infinite medium for synaptic inputs at different locations along the cell (once these make sense, #150 should be a cakewalk)

@stephanie-r-jones @jasmainak @rythorpe @ntolley Are you aware of existing code related to the above? I presume some must have been written for the calcium-investigations.

Not all of the above may belong into hnn-core, we should discuss it.

[jasmainak]

Part of my vision is to make the Cell object accessible with a public API. Then you can use Neuron API to do these things without having to bake them into HNN-core ... or to have thin wrappers if there are issues with respect to pickling etc. Have you looked at LFPy code? I think LFPy approach is "overkill" in the sense that they try to store everything which may consume too much memory.

[cjayb]

I really like the way Cell is materialising! I'm imagining what you describe: NEURON API calls with possible thin class-overrides to do things like virtual intracellular electrodes.

One goal would be to keep everything within reach: having to install LFPy and conform to yet another API seems overkill. We can just wait and see if anything remarkable comes out of this and decide then whether to include it in core.

In any case, I find doing these analyses necessary for implementing LFP electrodes, and would like to know if code already exists for any of the points above?

[rythorpe]

@cjayb I'm not aware of any code for the above mentioned items. Sounds like a worthwhile cause though.

[ntolley]

From a brief conversation with @stephanie-r-jones, I believe these are the intended behavior of the cells with a tonic current injection. Definitely correct me if I'm wrong!

L2 Pyr: adapting
L5 Pyr: bursting (with calcium)
Basket: regular firing