How much physics & EE for computational neuroscience?

[DreCate]

So there seems to have been many posts about how much math is required for computational neuroscience, but I don't know if anyone has raised the question of how much physics&EE is required for this field. It seems like circuit analysis is used quite extensively in modeling single neurons, but how much EE is actually needed to understand the literature? As for physics, I've seen something like Fick's law of diffusion appearing in some comp neuro and cellular physiology textbooks, but I don't know overall how much physics background is needed. So is a one-year college-level introductory physics adequate? Or is it necessary to study physics at a more advanced level (like a semester of electrodynamics and thermal/statistical physics)? Also, what textbooks would you recommend for physics&EE for comp neuro? Any advice is much appreciated.

 

[jbrussell93]

I can't say much as I'm still an undergrad as well, but I did take a Neurophysiology class last semester. I don't know about computational neuroscience but neurophysiology certainly makes use of plenty of EE and physics. The professor teaching the class was a neurophysiologist with an EE background so it may differ. The neuro concepts are pretty basic: diffusion relations, concentration gradients, etc but things have to be measured somehow. Current clamping and voltage clamping can be done in different ways with different circuit configurations that can get very complex. Also, look into dynamic clamping if you haven't heard of that. It is a fascinating subject at the moment and a great way to combine the computational side of things with the experimental. The idea is to hook the cell up to a computer and have the computer "insert" (virtually) ion channels into the cell by injecting the appropriate currents in real time. You can even run a simulation model through the cell.