Re: analysis of overlap/segregation betwee two colors in confocal image data

From: David Coder (dcoder@u.washington.edu)
Date: Mon Apr 17 2000 - 17:01:57 EST


Interestingly, the problem was answered by my next email (see below), but you
still want to know just where various classes of pixels are distributed
spatially--i.e., where the pixels are associated with cells and their particular
subcellular structures, or where within a cell are there regions of antigen
expression or coexpression.

Dave
dcoder@u.washington.edu

Date: Sun, 16 Apr 2000 22:56:35 +0100
From: Richard Herd <rahe@unixa.nerc-keyworth.ac.uk>
To: cyto-inbox
Subject: Re: ratio - 2 colour
Message-Id: <l03130300b51fe62c94b4@[209.88.233.113]>
Content-Type: text/plain; charset="us-ascii"

>> In addition, I'm
>>interested in plotting the intensities of a given region of pixels, to
>>compare the Red and green channels pixel by pixel (ie creating a
>>'colocalization graph' if plotted on the x and y axes). Im interested in
>>doing this either on binary (thresholded) images, or on 256 grey scale.


I did this a while ago and it was available on the ftp site
[http://rsb.info.nih.gov/nih-image/].  I don't think
it's still there, but I can prepare you a version of NIH-Image which will
let  you plot the XY scatter of pixel intensities for a pair of images in a
stack.  Alternatively, the source code if you want to put it in your
version.

Regards - Richard



Richard Herd
Montserrat Volcano Observatory
St Johns
Montserrat
West Indies
email : rahe@ua.nkw.ac.uk
phone : (1) 664 491 5647
FAX : (1) 664 491 2324


----- Original Message -----
From: Michael Dustin <dustin@pathbox.wustl.edu>
To: cyto-inbox
Sent: Friday, April 14, 2000 4:23 PM
Subject: analysis of overlap/segregation betwee two colors in confocal image
data



We are doing multiple color imaging of molecules involved in T cell
activation.  We need to develop or employ an existing index for the degree
of overlap (or segregation) between two different molecules that are labeled
with different dyes and imaged with a confocal microscope.  One idea would
be to start with a scatter plot (FL1 vs FL2) of the type used for two color
flow
cytometry where individual pixels would be plotted to define regions with
high overlap (on diagonal lower left to upper right) or areas with low
overlap (off diagonal).  We can generate lists of pixel values, but need to
translate the text files to a form readable by flow cytometry software like
Cell-quest to get the plots.  Is there software to go from Excel files to
the FCS format?  Has this problem been dealt with differently by others?
Thanks for any advice.

Michael L. Dustin, Ph.D.
Associate Professor of Pathology
Washington University School of Medicine
660 S. Euclid Ave
Campus Box 8118
St. Louis, MO 63110
Office- (314) 362-9618
Lab- (314) 362-8719
Fax- (314) 362-8888
Jerri Smith, Admin Assist. (314) 362-8740



This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:55:44 EST