q10 for Kinetic Scheme based channels

[borismarin]

Neuron (mod) generation for KS based channels does not honour q10 corrections.
This is due to rateScale being generated correctly, but not being applied to alphas/betas.

[borismarin]

Probably, the best solution would be somehow multiplying the rates here
by rateScale (a naive forwardRate="rf*rateScale" reverseRate="rr*rateScale" won't work because KineticScheme takes rr, rf from the transitions, which don't see rateScale)

[sanjayankur31]

@borismarin I expect this is still an issue?

I've been working on converting Zang et al 2018's purkinje cell model and there are a number of KS channels there. I haven't reached the part where I simulate the cell with NEURON, but I expect I'll hit this bug when I get there 🤔

[borismarin]

@sanjayankur31 I don't recall fixing it, but it seems you will be hitting it. What do you think about the proposed solution above? The Zang channels might be a good test.

[sanjayankur31]

Yeh, it'll have to be something like that. I'll discuss it with Padraig in the office next week and try to do a quick fix, since this will require a new release to the standard and APIs for it to be usable.

The two KS files from Zang et al 2018 are here:

• https://github.com/sanjayankur31/243446/blob/feat/neuroml-conversion/NeuroML2/channels/NaRSG.channel.nml
• https://github.com/sanjayankur31/243446/blob/feat/neuroml-conversion/NeuroML2/channels/SK2.channel.nml

I tested them out in jNeuroML and they replicate the original NEURON mod implementation but I haven't yet run them in generated NEURON:

https://github.com/sanjayankur31/243446/blob/feat/neuroml-conversion/NeuroML2/channels/20240604170737_test_narsg_states_NEURON.png vs https://github.com/sanjayankur31/243446/blob/feat/neuroml-conversion/NeuroML2/channels/20240607161612_test_narsg_states_NML.png

(test script here)

[sanjayankur31]

I think I maybe OK for these because I explicitly to the Q10 multiplication when defining the new component types for forward/reverse transitions' rates. (If I'd used a pre-existing rate component type from the standard, we'd have seen the issue?)