rallbranch

[thats_karlo]

Hello,

i found a function "rallbranch" in the example file of NEURON.like

create A
A{L=100 diam= 2 insert hh ralbranch=5}


I did not find any explanation or definitions for this function! could you help me to figure out what is that?

Thanks in advance,
karlo

[thats_karlo]

in this address you can find the examples for "rallbranch":

C:\nrn60\examples\nrniv\nrnoc

then look at " pascab3.hoc or pascab5.hoc or pascab5.hoc "

i can see the effect of this function, but need more information!

[hines]

rallbranch - 1 refers to the number of "virtual" copies of a section and its subtree (including all synapses) that connect to the parent. These copies take no time to compute (since they don't really exist except in the sense that the axial current flowing into the "real" subtree from the parent is 1/rallbranch smaller and the axial current flowing into the parent from the real subtree is rallbranch larger.
If you have
create d0, d1, d2
connect d1(0), d0(1)
connect d2(0), d1(1)
d1.rallbranch = 5
d2.rallbranch = 6
Then the simulation should be identical to
create d0, d1[5], d2[5][6]
for i=0,4 {
connect d1(0), d0(1)
for j=0, 5
connect d2[j](0), d1(1)
}
}

you just have to make sure that if you stimulate d2 in the first sim you stimulate all 30 d2[j] in hte second sim in exactly the same way.

Just make sure you test this whenever you use it. I imagine surprising results can occur if you connect LinearMechanism circuits to sections having a rallbranch != 1.
In fact I can't remember if I took into account all the edge cases involving
LONGITUDINAL_DIFFUSION, LinearMechanism, extracellular, etc.

[thats_karlo]

thank you so much, Hines!

it seems, there is no relation between "rallbranch" and Rall’s power rule,right?

[hines]

Well, I view rallbranch as a generalization of the rall 3/2 power rule in the sense that a cable with uniform diam sees the same impedance as if there is an abrupt change in diam of diam *= rallbranch^(-2/3)

[thats_karlo]

Thanks Hins, but i confused!!!!!

you said

"...rallbranch as a generalization of the rall 3/2 power rule...."

"rallbranch" create "virtual" copies of a section,

but when we create virtual sections, do their diameters hold Rall's power rule? if yes, then the example you gave above are not equal!

and if not, what is the advantage of "rallbranch" function? where and why we can use it?

[ted]

The "rallbranch" feature is principally useful for certain obscure programming tricks that
have nothing to do with equivalent cylinders per se--see
http://www.neuron.yale.edu/phpBB2/viewtopic.php?t=850
It should probably be avoided in any new models, not least because there is strong
empirical evidence that it is not obeyed by any real cells.

Comments about the "3/2 power rule" and the idea of an "equivalent cylinder"

The "3/2 power rule" is one of the three prerequisites that must be satisfied for a dendritic
tree to be reduced to an equivalent cylinder. The other two are:
diameters must be cylindrical
all dendritic terminations must be electrically equidistant from the soma
Quantitative morphometry proved long ago that none of these three assumptions is
satisfied by any real cell. But here we are, still talking about equivalent cylinders ~ 50 years
after the notion of an equivalent cylinder was first proposed, even though we now have
more than 20 years worth of detailed anatomical data that shows there aren't any cells
whose dendritic trees can be represented by equivalent cylinders.