ENH/VIZ: Introduce visualization of “currentscape”

[asoplata]

In biophysical modeling for computational neuroscience, understanding how the currents of individual channels contribute to the overall cell’s voltage fluctuations can be very helpful. One way to have per-cell visualization of the “currentscape” is defined by this paper: https://elifesciences.org/articles/42722 , where it’s a plot that illustrates the relative proportion of each current to a cell’s overall current:

This would be nice to have, and fortunately there is pre-existing code:

Pre-existing implementations:

• Note that the paper already includes some supplemental code for the vis: https://datadryad.org/stash/dataset/doi:10.5061/dryad.d0779mb#citations
• BlueBrain has also created a Python package of the same code that may be easier to bring in: https://github.com/BlueBrain/Currentscape

[dikshant182004]

i just wanna confirm that do i need to use the python package of BlueBrain to implement this or need to use that as a reference and figure out to how to implement it from the scratch

[asoplata]

Hey @dikshant182004 for new contributors, we prefer that they only work on one Issue/PR at a time and wait until that is completed, before moving on to others.

[dikshant182004]

now can i work on this issue

[asoplata]

Yes, I have merged your other PR #994, so feel free to try at this. I recommend trying to use the BlueBrain package first. However, depending on the complexity of the API, we may want to only use parts of their package directly, and simplify other parts. Please note that I removed the "good first issue" tag from this issue, since I expect it to be more complex than #994.

[dikshant182004]

hii @asoplata ,how i can access the individual channel currents ,because the output i am getting is the net current .since i need to get individual current to specify in config parameters .

[asoplata]

@ntolley @katduecker Is there a canonical way for a user to record per-mechanism current values for different sections or cells in HNN-Core? For example, can we record the ik of https://github.com/jonescompneurolab/hnn-core/blob/master/hnn_core/mod/km.mod#L92 separately from the net current of a particular section? Is it even currently possible to get such recordings?

There has definitely been some interest in recording this type of data in the past: #199 (comment) .

If we don't currently support this fine of a degree of recording, then this current issue will have to be put on hold. I think we should consider adding this functionality in the future, but it would certainly have to be a separate issue. Actually recording the data may be possible by expanding https://github.com/jonescompneurolab/hnn-core/blob/master/hnn_core/cell.py#L790-L799 , but considerable care would have to be taken with the necessary expansion of the output data structs and format.

[dikshant182004]

i have tried to modify the neuron model to record individual channel currents but fails to do so ,do tell me if i can help something in adding this funtionality .

[ntolley]

@asoplata at the moment this is not possible, the only supported ones are synaptic currents and calcium currents.

I definitely agree this could be useful a feature, I think I refrained from adding it early on because there wasn't any active projects that required this sort of recording, and none of the users were asking for it.