Hello!
I am new to the neuron simulation environment
I am currently having difficulty understanding the FH program on modelDB.
There seems to be some discrepancies between the leakage reversal potential in the actual paper and the program in modelDB.
The discrepancy seems to be quite large with the paper indicating that the reversal potential should be 0.026mV and the modelDB program using the value 69.74mV.
What’s interesting is that even with such a discrepancy the modelDB program outputs the same graphs as those in the paper.
I have also tried simulating the equations in matlab but to no avail as I'm getting stability issues even with robust algorithms such as ODE15s. The gating variables don’t seem to want to turn to the correct directions.
Any help would be greatly appreciated.
Matlab Simulation Graph:
FH ModelDB:
http://senselab.med.yale.edu/modeldb/Sh ... model=3507
FH Original Paper:
http://jp.physoc.org/content/171/2/302.full.pdf
Understanding Frankenhaeuser Huxley (FH)
Moderator: tom_morse

 Site Admin
 Posts: 5188
 Joined: Wed May 18, 2005 4:50 pm
 Location: Yale University School of Medicine
 Contact:
Re: Understanding Frankenhaeuser Huxley (FH)
Good question, and good to see that you have been actively investigating this. The answer is almost at hand. What variable is plotted in F&H Fig. 1, and what variable is plotted in the graph generated by NEURON that shows the spike? What initial value did F&H assume for V in their model (see p. 306), and what was the initial value of v in the NEURON implementation of their model? If you read the papers by Hodgkin and Huxley, do you recall how they dealt with the fact that spikes arise from a nonzero resting potential? The leakage reversal potential discrepancy contains a hint.aupi wrote:discrepancies between the leakage reversal potential in the actual paper and the program in modelDB.
The discrepancy seems to be quite large with the paper indicating that the reversal potential should be 0.026mV and the modelDB program using the value 69.74mV.
What’s interesting is that even with such a discrepancy the modelDB program outputs the same graphs as those in the paper.