Gap junction implementation based on average population membrane voltage

[rowleya]

A possible implementation of the "gap junction" behaviour would be to take the average membrane potential over time e.g. over 1s of all the neurons in a population, rather than individual neurons. This can then be sent to other populations once per second to compute the gap-junction current contribution input into the neurons.

[apdavison]

Can you explain your thinking here? Gap junction transmission is very rapid (0.1 ms) and very local, so I don't think averaging over time and over a population would be a good model.

[rowleya]

This was a user request locally - they were happy to use this as an approximation. I had assumed that the data transfer was fast, but they suggested this as a potential solution from their perspective. Clearly this is a specific implementation rather than a complete solution, but I said that I would put it in the issue list to make sure we don't forget it.

[andrewgait]

Leaving this issue here for future consideration.