Colocalization ROI

Hi Colocalization expert,

I’ve written a little script in Python to do some colocalization analysis measuring the Pearson and Spearman coefficient.
Unfortunately the datasets I am working with doesn’t have a separate channel to create the ROI and I’m afraid that using automatic threshold on one channel will bias the result. My questions are:

  1. Is it better in that case to use a fix threshold?
  2. Or alternatively, since for every images, there is only one cell in the middle of the field of view, I can create a bounding box area around the threshold cells.

My problems with the second solution is that it includes a lot of the background which “artificially” increase the coefficient.

Thanks,

Hi

You have to set the roi to the biologically relevant area where you want to measure. If you include background or other areas where the biology if interest is not present the correlation measurements will be skewed by that. So you need another data channel that allows you to define where the I tesesting biology is and is not

Else try an object based method

1 Like