From Wikipedia, the free encyclopedia

[edit] "obsolete" vs "incomplete" assumptions of synapse function

Reading the paper, I'd go with "obsolete". The new findings of Naundorf et al. suggest a difference in quality, not in quantity (context-sensitive differential regulation vs. a simple all-or-nothing concept). The title basically say it all - for the vertebrate (mammalian?) cerebral cortext, Hodgkin-Huxley dynamics is a behavior that can as it seems only be achieved in vitro and does not occur in a natural state. Note that not all models are "obsolete" unless proven obsolete, but rather preliminary or tentative. Dysmorodrepanis 14:57, 9 June 2006 (UTC)

This paper is interesting, but saying that it invalidates the Hodgkin-Huxley model is really kind of off-base. The HH model should be viewed as basically an inter-spike interval model -- it was generated as a model of when a neuron would throw a spike that had more biochemical plausibility than a simple leaky integrate-and-fire neuron. The temporal response sharpening the article discusses an is interesting example of the general phenomenon of cortical response nonlinearities; it probably does point out something useful about one of the many things that the HH model either doesn't or can't replicate. But at this point there's no widely accepted biochemical model of spiking behavior that doesn't describe the same basic mechanisms and time-course as the HH model. saccade 10, June 2006.

The point is that the paper suggests cortical neurons work differently, and such behavior seems crucial in understanding what exactly happens up there (nonlinear behavior seems the key; at any rate, our understanding of cerebral function seems to run into a barrier when viewed from a linear standpoint). Suffice to say that in mammalian cerebral cortices (as opposed to squid giant axons), the HH model doesn't really seem to be a particular adequate description. Dysmorodrepanis 12:08, 11 June 2006 (UTC)

It does raise an intriguing point, but it doesn't abandon most of the machinery of the HH model -- it suggests a change in the behavior of voltage gated ion channels, but doesn't question that they exist. Granted, the HH model does imply statistically independent stochastic gate behavior, which they are suggesting might not be a good assumption in cortical neurons. But bear in mind that most of the molecular machinery implied by the HH model has actually been found and independently observed -- it's a very well-validated model as far as what it was intended to model goes. Anyway, until the consensus in the molecular neuroscience community is that the HH model isn't really useful anymore, I think it would be kind of premature for Wikipedia to declare it obsolete. I know people who still do actual work with it, so I doubt it's on its way out just yet. saccade 11, June 2006.

Okay, to get technical on you guys, the original HH model is of type II excitability, where spikes occur through a Hopf bifurcation, while cortical APs are type I excitable, so spikes occur through a saddle-node bifurcation. So we know that the original HH model is not an accurate model of spike generation in cortical neurons, and no one would claim that it is. However, modified versions (that include A-type potassium currents) are believed to be accurate. Dysmorodrepanis, all of these models are fundamentally nonlinear, including the original HH model. The Naundorf et al. paper that claims to show cooperativity between Na+ channels is a controversial result, and in my opinion is likely an artifact. IIRC, the same results can be obtained using a standard HH model, if one takes into account the action potential being initiated at axon initial segment, which has been experimentally observed (I don't have the refs offhand). I don't know the full details, but I know of a group that will be publishing a rebuttal of Naundorf et al. soon (as of Dec. 2006). In any case, it's a recent paper, and it's a controversial result, so I don't think it belongs here in Wikipedia. evilrobotxoxo Dec 19, 2006.

So I actually have a reference to back this up-- http://jacknife.med.yale.edu/spikeclub/mc_resp.pdf -- in the mean time, I removed the references to the Naundorf paper entirely.Evilrobotxoxo 19:51, 19 December 2006 (UTC)

[edit] Link to FitzHugh-Nagumo model

I added a link to the FitzHugh-Nagumo model, this article does not exist yet in the English wikipedia. A German article is available at http://de.wikipedia.org/wiki/FitzHugh-Nagumo-Modell.

--193.175.8.13 13:13, 28 November 2006 (UTC)

[edit] a few problems with this article

This article has a few things that are not clearly explained and a few mistakes. I'm pretty sure the role of the Na+/K+ transporter was not modeled by H and H as a current. Basically, it was only reflected in the reversal potentials of the different conductances. Also, this article should explicitly mention what the main conductances were, etc. I'll make some changes when I get a chance, but I wanted to notify first. Evilrobotxoxo 03:54, 20 December 2006 (UTC)