Re: colocalization analysis

Posted by Judy Trogadis on
URL: http://confocal-microscopy-list.275.s1.nabble.com/colocalization-analysis-tp786850p837000.html

Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Dear Julio

You are wonderful, I really appreciate your taking the time to think about this problem and then to discuss it in such depth. Your comments are very helpful and have given me fodder for thought. As is often the case with a seemingly simple question, the answer invariable begins with . . . . "it all depends on ". . . .  

Armed with this information and comments from others, I will probably give a presentation to the users about colocalization analysis.
Thank you.
Judy


>>> Julio Vazquez <[hidden email]> 09/03/08 1:59 PM >>>
Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal 

Judy,

Just a few more thoughts:

There have been a number of papers on colocalization. I just found a  
recent one in Current Protocols in Cell Biology that gives a working  
protocol so to say for doing colocalization analysis with the ImageJ  
colocalization plugins:

Zinchuck and Zinchuk, Quantitative Colocalization Analysis of  
Confocal Fluorescence Microscopy Images. CPCB 4.19.1-4.19.16, June 2008.

This can be obtained from the Journal's web site.


The various colocalization coefficients are derived from the  
Correlation coefficients used in statistics.  Wikipedia defines  
correlation as a measure of the relationship between two variables.  
Maybe the various correlation coefficients could be modified for use  
with three variables, but I am not even sure how the numbers could be  
interpreted, since different combinations of intensities could  
probably give similar results, and a lot of potentially useful  
information would be lost. So a workaround when using this type of  
approach would be to do pairwise colocalization comparisons (A vs B,  
B vs C and C vs A).

If my concern were to evaluate the presence of three proteins (or  
other markers) inside a cell, my approach would be to select suitable  
thresholds for each signal, generate a mask for each channel with the  
selected thresholds, and then combine the three masks through Boolean  
or Image Arithmetic operations. This would generate a channel where  
all three markers are present (above set threshold). You could also  
create masks by combining two markers, and use that mask to analyze  
the distribution of the third marker. I would then quantitate the  
different markers through those masks (integrated intensity, surface  
area/volume, or whatever variable is the most relevant biologically),  
and either look at the absolute numbers, or the relative numbers as a  
function of the total cell area, or total intensity in any given  
channel, etc... To be even more accurate, I could make a mask for the  
cells (or nuclei, etc...) and restrict my analysis to those regions.  
As I mentioned in the previous post, it should be up to the  
investigator to let you know which of these numbers or pairwise  
comparisons are the most relevant biologically for their experiment.  
With the mask approach you can determine, for instance, that 20% of  
protein A is present in regions where proteins B and C are also  
present, and so on... this is very easy for me to grasp intuitively.  
I will also be able to know how large the colocalized patches are,  
how many there are, where they are located, and so on... In my  
opinion, this is much more informative that a Pearson's coefficient  
of, let's say, + 0.43.

My main issue with the standard correlation approach is that it is  
pixel-based, as opposed to object-based, and that the results are  
highly dependent on how the images were collected (sampling) and  
processed, how much noise and background they have, how well  
registered the different channels are (always true, but even more so  
in this case, especially if looking at colocalization of very small  
objects) and which regions and thresholds are used for analysis. Used  
properly, these methods are quite powerful though, and will give very  
useful information such as colocalization coefficients,  % of A  
colocalized, % of B colocalized, etc....  but I have seen so many  
results that didn't make sense just because someone was using a  
poorly sampled image, or a noisy image (such as a confocal image  
under limiting signal conditions), that I am very careful and tend to  
prefer a more visual approach where objects are identified based on  
thresholds, and then analyzed.  Anyhow, the paper above discusses  
some of the issues to take into consideration when performing the  
pixel-based type of colocalization analysis.

hope this helps,

with kind regards,

Julio.


--
Julio Vazquez
Fred Hutchinson Cancer Research Center
Seattle, WA 98109-1024

http://www.fhcrc.org/ 

-

On Aug 27, 2008, at 6:58 AM, Judy Trogadis wrote:

> Search the CONFOCAL archive at
> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal 
>
> Hi, Michael,
>
> You're right, 2 runs would do it - but one of the users of our  
> imaging facility has triple labelled preps ready to analyze. He is  
> looking for the presence of 3 proteins in a cell but I am not sure  
> about their proximity to each other. Visual observation at high  
> magnification could give a clue. I think the user wants some  
> numerical value for a grant.
>
> Thanks.
> Judy
>
>
>>>> Michael Weber <[hidden email]> 08/27/08 9:45 AM >>>
> Search the CONFOCAL archive at
> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal 
>
> Judy,
>
> colocalization analysis is quite straight forward with ImageJ and the
> "Colocalisation Threshold" plus "Colocalization Test" plugin  
> according to
> Costes.
>
> Regarding triple labeling, which type of questions do you plan to  
> answer?
> I can think about a scenario with two marked structures and how  
> they are
> colocalizing with the nuclei - then you do two runs: nuclei vs.  
> staining
> 1, nuclei vs. staining 2. I am not aware of an established three-color
> colocalization equation. Or do I miss something here?
>
> Michael
>
>
>> Search the CONFOCAL archive at
>> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal 
>>
>> I have a triple labelled sample and would like to do colocalization
>> analysis. What approaches are most people using? The plugins I  
>> have seen
>> or used only handle double labelled specimen. A 3-D fluorogram  
>> perhaps?
>>
>> Thank you.
>>
>> Judy Trogadis
>> Bio-Imaging Coordinator
>> St. Michael's Hospital, 7Queen
>> 30 Bond St.
>> Toronto, ON M5B 1W8, Canada
>> ph:  416-864-6060  x6337
>> pager: 416-685-9219
>> fax: 416-864-6043
>> [hidden email]