Typesetting error in nap.mod affecting multiple ModelDB entries

Discussions of particular models.

Moderator: tom_morse

Post Reply
Danko Georgiev
Posts: 8
Joined: Fri Oct 15, 2010 10:41 am
Location: Varna, Bulgaria
Contact:

Typesetting error in nap.mod affecting multiple ModelDB entries

Post by Danko Georgiev »

The nap.mod file, which is reused in multiple models, contains an error in the n_inf formula.

Below, I list 10 ModelDB entries that can be accesed through the links and the reused links:

Error in 1 ModelDB reused in 4 other ModelDB entries

https://modeldb.science/showmodel?tab=2 ... nt/nap.mod

The incorrect code is

Code: Select all

n_inf = 1 / (1 + (exp(vhalf - v)/K))
The correct code based on the Supplementary Material from Poirazi P, et al. Neuron 2003; 37 (6): 977-987 should be:

Code: Select all

n_inf = 1 / (1 + exp((vhalf - v)/K))
The incorrect placement of the opening bracket before exp leads to both shift in apparent vhalf < - 50.4 mV and also effectively sets K = 1 mV.

The intended correct values are

Code: Select all

vhalf	= -50.4 (mV)
K	= 4.5 (mV)
Error in 1 ModelDB not reused

https://modeldb.science/showmodel?tab=2 ... ap/nap.mod

The incorrect code is

Code: Select all

n = 1 / (1 + (exp(vhalf - v)/K)) : steady state value
Error in 1 ModelDB reused in 2 other ModelDB entries

https://modeldb.science/showmodel?tab=2 ... sm/nap.mod

The incorrect code again is

Code: Select all

n = 1 / (1 + (exp(vhalf - v)/K)) : steady state value
Again, Error in 1 ModelDB not reused

https://modeldb.science/showmodel?tab=2 ... 15/nap.mod

One instance of the error has been spotted by Ted in the NEURON forum

viewtopic.php?t=86

Code: Select all

n = 1 / (1 + (exp(vhalf - v)/K)) : steady state value
The error propagates outside of ModelDB into the ICGenealogy, CNCB, University of Oxford, which apparently plot the incorrect nap.mod channel

https://icg.neurotheory.ox.ac.uk/channels/2/1668

I think that this error should be flagged across affected ModelDB entries.

To identify the true number of affected ModelDB entries, I would suggest internal search through all .mod files of the ModelDB database of the erroneous piece of code

Code: Select all

(exp(vhalf - v)/K)
because I was only relying on external Google search to find affected ModelDB entries.
Post Reply