How to move neuron in space without changing its dimensions

[NanTian1984]

Hei,

I'm a beginner in neuron. I have been trying to move preconstructed neuron cell in 3D but have little luck of this task. Whenever I'm initiating (pt3dadd) or moving(pt3dchange) the soma (or anyother section of the neuron), NEURON changes the lenght of the segment to fit the coordinate I've put in. I'm wondering is there a embeded function in NERUON that would allow me to move the segment without changing its demension. Do you think I should change the model code a little to make the model dynamic?

Plus, I'm still a beginner. Is there a simple tutorial that is plain about how to move the cell in NEURON, since I felt Andrew Gillies and David Sterratt's tutorial is not detailed enough on topic of moving cells.

Thanks a lot.

-Nan

[ted]

The root* section controls the apparent location of a model cell in a
Shape plot. The easiest way to control the apparent location of a model
cell is to create a small dummy section that is used as the root. Then
you won't have to tinker with any of the pt3dadd statements for any
of the sections you care about.

So if the root of your model is called somesection, and you
want the model to appear to originate at
XORIGIN, YORIGIN, ZORIGIN
all you have to do is add the statements

Code: Select all

XORIGIN = your_x_value 
YORIGIN = your_y_value
ZORIGIN = your_z_value
create dummy
dummy {
  pt3dadd(XORIGIN-0.1, YORIGIN-0.1, ZORIGIN-0.1, 0.2)
  pt3dadd(XORIGIN+0.1, YORIGIN+0.1, ZORIGIN+0.1, 0.2)
}
connect somesection(0), dummy(0.5)

before anything that creates a Shape plot. A good place to insert
this would be right after the code that defines the structure of your
model cell.

dummy will be too small to have a significant effect on any model
that has reasonable membrane properties and a surface area larger
than 10 um2, and you will have complete control over the apparent
location of the model cell without jeopardizing its geometry.

*--The root of a tree is the branch which has no parent.

[ssothro]

Apart from moving a reconstructed cell is there an easy way to rotate it in 3D space?

Thanks

[ted]

Apart from moving a reconstructed cell is there an easy way to rotate it in 3D space?

You can view it from different angles. See
http://www.neuron.yale.edu/neuron/stati ... /shape.htm
which is from this tutorial on how to analyze the electrotonic properties of a cell
http://www.neuron.yale.edu/neuron/stati ... zclass.htm

[gpho]

Is there a way to actually move or rotate a neuron in space?

<< Moderator's apology: Sorry, somehow I started editing your post instead of answering it.
--ted >>

[ted]

You could apply a rotation transformation to all pt3d points. pt3dchange is the hoc procedure that lets you apply the change. Just add your own rotation matrix and a bit of code to iterate over all pt3d points of all sections, and stir.

I think I'd stick to changing the orientation of the field . . .