Wiki » History » Version 4
Version 3 (Boris Marin, 30 Apr 2014 14:58) → Version 4/7 (Boris Marin, 30 Apr 2014 14:58)
L5 Pyramidal Neuron Almog and Korngreen (2014)
==============================================
Model description
-----------------
This model originally appeared in **Almog \*Almog M, Korngreen A \* (2014)\* A Quantitative Description of Dendritic Conductances and Its Application to Dendritic Excitation in Layer 5 Pyramidal Neurons [J Neurosci 34(1):1](http://www.jneurosci.org/content/34/1/182).
The relevant *neuron* files can be downloaded from [modelDB](http://senselab.med.yale.edu/modeldb/ShowModel.asp?model=151825).
h2.
Conversion to NeuroML
---------------------
We are currently (as of Jan/2014) converting the original model to NeuroML. Most *.mod* mechanisms can be mapped to [ChannelML](http://neuroml.org/NeuroMLValidator/Transform.jsp?localFile=NeuroMLFiles/Schemata/v1.8.1/Level2/ChannelML_v1.8.1.xsd&xslFile=NeuroMLFiles/Schemata/XSD_Readable.xsl) with exception of the Calcium mechanisms described in terms of the [GHK](http://en.wikipedia.org/wiki/GHK_flux_equation) formalism. These latter needed to be translated to [NeuroML2](http://www.neuroml.org/neuroml2.php), which supports arbitrary current laws.
h3.
### *NeuroConstruct* project
h4.
#### Ion Mechanisms
Both original (*.mod*) and NeuroML versions of the ionic mechanisms were imported into a *neuroConstruct* project, with simulation configurations allowing the comparison of both implementations. Mechanisms were named according the following convention: the original *.mod*\_ mechanisms were suffixed with **nrn*, while the NeuroML mechanisms had no suffix.
h4. Morphology
Cell [morphology](http://senselab.med.yale.edu/modeldb/ShowModel.asp?model=151825&file=\Demo\cells\A140612.hoc) from the*modelDB\_ project was exported to NeuroML using *neuron*.
h3.
### Idiosyncrasies in the conversion
h4. Inhomogeneous parameters over the dendritic tree
p=. ![]({width: 80%}gih_comparison.png)
The original *neuron* version results in piecewise linear interpolation of the distribution over each section.
Compare the *neuroConstuct* generated code
\<pre\>
//neuroconstruct generated
objref PathLengthApicalDends
PathLengthApicalDends = new SubsetDomainIterator
PathLengthApicalDends.update {
x = PathLengthApicalDends.x
p = PathLengthApicalDends.p
gmax\_iA = 100000.0** (2.82824E-4 + (0.0033165 \* exp(((–1) \* 0.0117721) \* p))) // 100000.0 to convert from nc to NEURON units
}
</pre>
With the original *.hoc* specification
gbar_iA(0:1) = (gka_start+gka_beta*exp(gka_alpha*dist1)):(gka_start+gka_beta*exp(gka_alpha*dist2)) ![](gih_comparison.png)
==============================================
Model description
-----------------
This model originally appeared in **Almog \*Almog M, Korngreen A \* (2014)\* A Quantitative Description of Dendritic Conductances and Its Application to Dendritic Excitation in Layer 5 Pyramidal Neurons [J Neurosci 34(1):1](http://www.jneurosci.org/content/34/1/182).
The relevant *neuron* files can be downloaded from [modelDB](http://senselab.med.yale.edu/modeldb/ShowModel.asp?model=151825).
h2.
Conversion to NeuroML
---------------------
We are currently (as of Jan/2014) converting the original model to NeuroML. Most *.mod* mechanisms can be mapped to [ChannelML](http://neuroml.org/NeuroMLValidator/Transform.jsp?localFile=NeuroMLFiles/Schemata/v1.8.1/Level2/ChannelML_v1.8.1.xsd&xslFile=NeuroMLFiles/Schemata/XSD_Readable.xsl) with exception of the Calcium mechanisms described in terms of the [GHK](http://en.wikipedia.org/wiki/GHK_flux_equation) formalism. These latter needed to be translated to [NeuroML2](http://www.neuroml.org/neuroml2.php), which supports arbitrary current laws.
h3.
### *NeuroConstruct* project
h4.
#### Ion Mechanisms
Both original (*.mod*) and NeuroML versions of the ionic mechanisms were imported into a *neuroConstruct* project, with simulation configurations allowing the comparison of both implementations. Mechanisms were named according the following convention: the original *.mod*\_ mechanisms were suffixed with **nrn*, while the NeuroML mechanisms had no suffix.
h4. Morphology
Cell [morphology](http://senselab.med.yale.edu/modeldb/ShowModel.asp?model=151825&file=\Demo\cells\A140612.hoc) from the*modelDB\_ project was exported to NeuroML using *neuron*.
h3.
### Idiosyncrasies in the conversion
h4. Inhomogeneous parameters over the dendritic tree
p=. ![]({width: 80%}gih_comparison.png)
The original *neuron* version results in piecewise linear interpolation of the distribution over each section.
Compare the *neuroConstuct* generated code
\<pre\>
//neuroconstruct generated
objref PathLengthApicalDends
PathLengthApicalDends = new SubsetDomainIterator
PathLengthApicalDends.update {
x = PathLengthApicalDends.x
p = PathLengthApicalDends.p
gmax\_iA = 100000.0** (2.82824E-4 + (0.0033165 \* exp(((–1) \* 0.0117721) \* p))) // 100000.0 to convert from nc to NEURON units
}
</pre>
With the original *.hoc* specification
gbar_iA(0:1) = (gka_start+gka_beta*exp(gka_alpha*dist1)):(gka_start+gka_beta*exp(gka_alpha*dist2)) ![](gih_comparison.png)