Using Strahler's analysis to reduce realistic models (Marasco et al, 2013)

 Download zip file 
Help downloading and running models
Accession:149000
Building on our previous work (Marasco et al., (2012)), we present a general reduction method based on Strahler's analysis of neuron morphologies. We show that, without any fitting or tuning procedures, it is possible to map any morphologically and biophysically accurate neuron model into an equivalent reduced version. Using this method for Purkinje cells, we demonstrate how run times can be reduced up to 200-fold, while accurately taking into account the effects of arbitrarily located and activated synaptic inputs.
Reference:
1 . Marasco A, Limongiello A, Migliore M (2013) Using Strahler's analysis to reduce up to 200-fold the run time of realistic neuron models. Sci Rep 3:2934 [PubMed]
Model Information (Click on a link to find other models with that property)
Model Type: Neuron or other electrically excitable cell; Dendrite;
Brain Region(s)/Organism: Hippocampus; Cerebellum;
Cell Type(s): Hippocampus CA1 pyramidal GLU cell; Cerebellum Purkinje GABA cell;
Channel(s): I Na,t; I T low threshold; I K; I Calcium; Ca pump;
Gap Junctions:
Receptor(s): AMPA;
Gene(s):
Transmitter(s): Glutamate;
Simulation Environment: NEURON;
Model Concept(s): Activity Patterns; Active Dendrites; Influence of Dendritic Geometry; Detailed Neuronal Models; Action Potentials; Synaptic Integration;
Implementer(s): Limongiello, Alessandro [alessandro.limongiello at unina.it];
Search NeuronDB for information about:  Hippocampus CA1 pyramidal GLU cell; Cerebellum Purkinje GABA cell; AMPA; I Na,t; I T low threshold; I K; I Calcium; Ca pump; Glutamate;
/
PurkReductionOnLine
morphologies
readme.txt
CaE.mod *
CalciumP.mod *
CaP.mod *
CaP2.mod *
CaT.mod *
K2.mod *
K22.mod *
K23.mod *
KA.mod *
KC.mod *
KC2.mod *
KC3.mod *
KD.mod *
Kdr.mod *
Kh.mod *
Khh.mod *
KM.mod *
Leak.mod *
NaF.mod *
NaP.mod *
pj.mod
clusterisingMethods.hoc
fixnseg.hoc
mergingMethods.hoc
mosinit.hoc
ranstream.hoc *
RedPurk.hoc
stimulation1.hoc
useful&InitProc.hoc
                            
/* Sets nseg in each section to an odd value
   so that its segments are no longer than 
     d_lambda x the AC length constant
   at frequency freq in that section.

   Be sure to specify your own Ra and cm before calling geom_nseg()

   To understand why this works, 
   and the advantages of using an odd value for nseg,
   see  Hines, M.L. and Carnevale, N.T.
        NEURON: a tool for neuroscientists.
        The Neuroscientist 7:123-135, 2001.
*/

// these are reasonable values for most models
freq = 100      // Hz, frequency at which AC length constant will be computed
d_lambda = 0.1

func lambda_f() { local i, x1, x2, d1, d2, lam
        if (n3d() < 2) {
                //print "dcdc: ", diam, diam/(4*PI*$1*Ra*cm)
                return 1e5*sqrt(diam/(4*PI*$1*Ra*cm))
        }
// above was too inaccurate with large variation in 3d diameter
// so now we use all 3-d points to get a better approximate lambda
        x1 = arc3d(0)
        d1 = diam3d(0)
        lam = 0
        for i=1, n3d()-1 {
                x2 = arc3d(i)
                d2 = diam3d(i)
                lam += (x2 - x1)/sqrt(d1 + d2)
                x1 = x2   d1 = d2
        }
        //  length of the section in units of lambda
        lam *= sqrt(2) * 1e-5*sqrt(4*PI*$1*Ra*cm)

        return L/lam
}

proc geom_nseg() {
  soma area(0.5) // make sure diam reflects 3d points
  forall { nseg = int((L/(d_lambda*lambda_f(freq))+0.9)/2)*2 + 1  }
}


Loading data, please wait...