ROI problems with Colocalization (JACoP and Coloc2)

Hello there,

I am pretty new in using ImageJ and have encountered some problems when trying to do colocalization analysis.
I have stained the nucleus of my cells with DAPI and want to colocalize it to another dye (which is obviously not in the nucleus, I just need to prove it with some data.) As far as I know, Pearson’s coefficient should be a negative value, because it is anticolocalized.)
When I tried colocalization with JACoP, I got positive Pearson’s coefficients. I think this is because it takes the whole image with a lot of background into account and not the single cells, that I have selected as ROI.
Therefore, I switched to coloc2, which indeed gives me a negative Pearson’s coefficient.
Still, I struggle with the Mander’s coefficients: When I take my whole image with many cells on it and select one cell as ROI, I get m1 = 0.00 and a m2 = 1.00. (The Pearson’s coefficient is -0.5)
When I crop the image and clear everything outside the ROI, the m1 = 0.349 and the m2 = 0.296. (Pearson’s coefficient is -0.58)

-What is the reason for the strange Mander’s coefficients in the first case?
-Why do Pearson’s and Mander’s coefficients change, although I have selected the exact same ROI?

Thank you very much for your help, I appreciate it very much.

Hi @zech

Welcome to the Forum. It is a bit confusing to figure out what exactly is going on in your particular case. It would be great if you could post some example images. The best would be to post original datasets/ROIs and your step-by-step process - that would help us reproduce the discrepancy you are finding.

But without all that, I only have a hunch… if you crop the image and clear everything outside the ROI - the calculation will change if the ROI is still not ‘active’. So are you sure this is the case? Is there are reason for you cropping/clearing the image for each calculation?

Perhaps others will have better insight/advice.

eta