www.neuron.yale.edu

I am trying to replicate the NEURON model described in this paper:

A cell model study of calcium influx mechanism regulated by calcium dependent potassium channels in Purkinje cell dendrites
Koji Chono, Hiroshi Takagi, Shozo Koyama, Hideo Suzuki, Etsuro Ito

However, I just cannot do it. I have contacted the prinipal author, Etsuro Ito but he seems to be unreachable. Replication should be elementry because this model is just a variant of the following NEURON model - has same equations, but just with different parameter values:

http://senselab.med.yale.edu/senselab/m ... odel=17664

Maybe I am just making mistakes in my coding. Has anyone else tried this? Would be so, so grateful to anyone that might be able to help.

mikey wrote:I am trying to replicate the NEURON model described in this paper:
. . .
However, I just cannot do it.

What exactly is the manifestation of this failure to replicate?

I have contacted the prinipal author

Try one or more of the other authors.

model is just a variant of the following NEURON model - has same equations, but just with different parameter values

Typographical errors are common. Check the journal for notices of errata. Check the
authors' own WWW sites to see if they mention errata or have made the code available.

Maybe I am just making mistakes in my coding.

If the only differences are numeric values of parameters, what coding is necessary?

Q) What exactly is the manifestation of this failure to replicate?
A) My replicated model does not produce the same voltage-time relationships as depicted in the figures of the model. Using same stimulation protocal as paper.

Q) I have contacted the prinipal author Try one or more of the other authors.
A) I have tried to get contact details for them but alas...cannot. Also, the authors are Japanese and I think perhaps as you go along the authors - the chances of speaking english become less.

Q) Typographical errors are common. Check the journal for notices of errata.
A) Will do.

Q) Check the authors' own WWW sites to see if they mention errata or have made the code available.
A) Sadly not.

Q) Maybe I am just making mistakes in my coding.
If the only differences are numeric values of parameters, what coding is necessary?
A) True. No coding going on. Maybe I am making mistakes in my alteration of parameter values. But I think not - as it is a simple task and I have double checked a few times now.

Q) Typographical errors are common. Check the journal for notices of errata.
A) Sadly no errata that could clear this up for me.

One problem is that the equations are listed in the paper as:

m
alpha = A/B+exp((v+C)/(D))
beta = E/F+exp((v+C)/(D))
h
alpha = A/B+exp((v+G)/(H))
beta = E/F+exp((v+G)/(H))

I think this must be a mistake because I think it should be:

m
alpha = A/B+exp((v+C)/(D))
beta = E/F+exp((v+G)/(H))
h
alpha = A/B+exp((v+C)/(D))
beta = E/F+exp((v+G)/(H))

I have coded it up as I think it should be. Because Im thinking that the equations as they are in the paper must be wrong

??
If the model is identical to an entry in ModelDB, except for numeric values of parameters,
then the contents of the mod files in the ModelDB entry specify the form of the equations.