www.neuron.yale.edu

Hello. I would like input on which direction I should take. I would like to do do manipulations on BK channels in CA1 pyramidal neurons based on some electrophysiological experiments my lab has done/will do. I am a beginner at NEURON though I do have a strong background in programming. I see two outcomes. The first is finding a model "in the wild". The second is taking a model that is similar enough to what we want to do and add the extra functionality.

Regarding the first approach:
I started off looking at some literature a past labmate provided. However, neither Gu et al (2007) nor Siskova et al (2014) have source code publicly available and Siskova et al does not implement BK functionality in their model anyway. I also combed through the ModelDB database. I found some that were promising. However, I have yet to find a model that implements exactly what I would like (CA1 pyramidal neuron with BK channel functionality) and I tried using the native search function as well as google dorking.

Regarding the second approach:
I was able to find some models through the site that looked really close to what we wanted but weren't ideal: Almog and Korngreen (2014) (implements a neocortical pyramidal neuron with BK channel functionality), Bianchi et al (2012) (CA1 pyramidal neuron with no BK functionality). Shao et al (1999) would be ideal but I cannot seem to find source code online.

In the event I had to extend the native functionality of the model, I would like guidance on how to go about it. I have poured through source code from other neuronal types. Some implement BK channels in three states (open->inactivated->closed), some implement channels with two states (open->close) or even one. I have been searching for a good tutorial kind of like the one for the subthalamic neuron MIT has.

Thanks

Good questions.

Shao et al (1999) would be ideal but I cannot seem to find source code online.

You might ask the authors if they are willing to share it with you. Or maybe they would even be willing to deposit it in ModelDB. Note that after 1/6th of a century, many authors will not be able to find the code they used; many others will find N different versions of it but not be able to tell which version was used for any particular figure in the paper.

It would make sense to start with a reasonable model of a Ca1 pyramidal cell, add a BK mechanism to it, and then make whatever parameter adjustments are required so that the resulting model is a reasonable approximation to what you see experimentally. Both Bianchi et al (2012) and Almog and Korngreen (2014) involve modelers who have done good work.

I do have a strong background in programming

so you know that not all code is of equally high quality, and one must carefully examine and test any code to make sure it works as advertised.

Some implement BK channels in three states (open->inactivated->closed), some implement channels with two states (open->close) or even one.

Some know or care more about those details than others do. How you proceed should be governed by your intent. Are the detailed properties of BK channels of central importance to your hypothesis? If so, you'll want an implementation that preserves those details. If BK channels are on the "conceptual margins" of your hypothesis, i.e. you just need something that roughly approximates their dynamics and their dependence on voltage and calcium, then a simpler representation will suffice.

I have been searching for a good tutorial kind of like the one for the subthalamic neuron MIT has.

What one is that?

Thank you for your response.

Some know or care more about those details than others do. How you proceed should be governed by your intent. Are the detailed properties of BK channels of central importance to your hypothesis? If so, you'll want an implementation that preserves those details. If BK channels are on the "conceptual margins" of your hypothesis, i.e. you just need something that roughly approximates their dynamics and their dependence on voltage and calcium, then a simpler representation will suffice.

Yes I think a simpler implementation would suffice so long as it approximates the experimental data, like you stated.

Which one is that?{/quote]

Here is the MIT tutorial in question. It demonstrates how to make a simplified subthalamic neuron (and a neural network) from scratch.
http://web.mit.edu/neuron_v7.3/nrntutht ... /tutA.html
http://web.mit.edu/neuron_v7.3/nrntutht ... /tutB.html
http://web.mit.edu/neuron_v7.3/nrntutht ... /tutC.html
http://web.mit.edu/neuron_v7.3/nrntutht ... /tutD.html
http://web.mit.edu/neuron_v7.3/nrntutht ... /tutE.html

Actually that's an old tutorial by Andrew Gillies and David Sterratt of Edinburgh University. NEURON's Documentation page http://www.neuron.yale.edu/neuron/docs has had a link to the original for a long time. You might want to take a look at some of the items on the Documentation page below the header Key papers about NEURON.