Hi Aaron,
I'm having to do the same thing more or less. My general approach is
to local either the pre or post-synaptic marker as my landmark. Let's
say I'm going to start with a presynaptic red marker. I locate the
center of mass witrh Adrian's FWHM plugin, then perform a radial sweep
of line scans around the center of mass of that land mark (presumably a
puncta of some size and shape) to look for local maxima in the green
"post-synaptic" channel/color. This radial sweep can easily be converted
into a 2D image where X & Y represent distance (from your puncta center)
and angle of the sweep. Obviously you are only going to look for puncta
(or local maxima) within a reasonable distance from the presynaptic
marker. I then use "Find local maxima" on this 2D projection of the
radial sweep to identify possible post-synaptic puncta. You can
geometrically calculate the position of this local max, and subject it
to whatever distance or intensity criteria you like, and if it passes
add it to a list of "neighboring" (i.e. not exactly colocalized) puncta
relative to your presynaptic puncta, which is presumably one of a long
list of puncta in your ROI manager. This macro is still under
development, but it seems to do an OK job.
Damon
Post doctoral fellow, Tsien Lab
Molecular & Cellular Physiology
Stanford University.
On 8/25/2010 11:58 AM, Aaron Johnson wrote:
> Dear listserv,
> I'm looking for suggestions on how to measure the relationship between the pre& post synaptic sides of a synapse. A "good" relationship would be like holding your hands together with the facing sides matching, while a "bad" relationship would be like if one of your hands were a fist and the other spread open.
>
> My data are 2-color confocal stacks with each side labeled with a different fluorophore. This seems like some sort of 3D colocalization problem, but not quite - the fluorophores aren't in the same location. It's the surface area of each facing side that I'm trying to compare. I'm envisioning binarizing the images, creating two objects and then somehow comparing the percentage of surface areas (in 3D!) that match, but I can't quite get my head around how to do this.
>
> Any help or direction would be much-appreciated.
>
> Thanks!
>
> Aaron
>
> -------------------------
> Aaron M. Johnson
> Pre-Doctoral Fellow
> Dept. of Surgery/Div. of Otolaryngology
> University of Wisconsin, Madison
Locked
