www.neuron.yale.edu

Hi,
I have a datafile for 3D reconstructed neuron (the tree is defined with "pt3dadd" commands). What I would like to do is to "clone" the specific branch (with all its sub-branches) of this given tree. That is, I want to create a new tree in which, at a specific point, there will be two exactly identical (morphology-wise) sub-branches.
My solution was just to go and look into the data file, find the desired branching point, and create the "clone" (by copy-pasting and changing the name of the new branch), and attach it at the same point as the original branch. Since the data is 3D, this means that, unless I change the 3D coordinates of the clone, when I try to visualize the new tree, the clone and the original branch will overlap.
My questions are:
1) following this procedure, even though the 3D locations of the two branches (the original one and the clone) are completely identical, the branches are treated as separate objects, right? I presumed that the 3D information is just for visualization purposes and does not affect the computation.
2) for that matter (easy visualization) is there any simple and nice way to switch from the 3D format to the "dimensionless" format .
Many thanks for the answer.

delarue wrote:What I would like to do is to "clone" the specific branch (with all its sub-branches) of this given tree. That is, I want to create a new tree in which, at a specific point, there will be two exactly identical (morphology-wise) sub-branches.

I thought this could be done with the CellBuilder, but now that I try it, I see that the cloned sections use stylized specification with default geometry (L 100, diam 500)--so topology is preserved, but pt3d data are not.

So, as tedious as it may seem, you're doing it the right way--

My solution was just to go and look into the data file, find the desired branching point, and create the "clone" (by copy-pasting and changing the name of the new branch), and attach it at the same point as the original branch. Since the data is 3D, this means that, unless I change the 3D coordinates of the clone, when I try to visualize the new tree, the clone and the original branch will overlap.

Just change the coords associated with the root section of your new subtree in a way that preserves distances between adjacent 3d measurement points. The simplest strategy is to invert the signs of the deltas along the x axis or the y axis or both. The offset of the cloned subtree will depend on the length and orientation of the original subtree's root section.

1) following this procedure, even though the 3D locations of the two branches (the original one and the clone) are completely identical, the branches are treated as separate objects, right?

True. Where things are drawn in a shape plot has nothing to do with the equations that NEURON solves.

2) for that matter (easy visualization) is there any simple and nice way to switch from the 3D format to the "dimensionless" format .

The term is "stylized rendering" and shape plots do it only for sections whose geometry is specified with the stylized style, i.e. in terms of L and diam. Shape plots of cells whose geometry is specified with the pt3d style preserve every kink and wiggle. The CellBuilder takes a shortcut of rendering sections as straight lines that connect the coords of their 0 and 1 ends, but I don't know if that's doable outside of the CellBuilder class.

ted wrote:

2) for that matter (easy visualization) is there any simple and nice way to switch from the 3D format to the "dimensionless" format .

. . .
The CellBuilder takes a shortcut of rendering sections as straight lines that connect the coords of their 0 and 1 ends, but I don't know if that's doable outside of the CellBuilder class.

On a Shape plot select the menu item ShapeStyle/Schematic