Double staining with DAPI colocalization analysis problems

I just started using CellProfiler yesterday and I’ve already learned a lot thanks to the tutorials and this forum.

I am trying to analyze IF cell staining with DAPI, CD45, and DsRed with the hopes of calculating single and double positives.

I think I have the pipeline working but I’m having trouble adjusting the parameters. Some of the objects look like they are more than one cell. I’m hoping you can point out what I’m doing wrong. I’m unfamiliar with the different thresholds to choose from.

Thanks,
Paige






12-0327 Paige.cp (20.2 KB)

Hi Paige,

Re: “objects look like they are more than one cell” - Assuming you are referring to the DAPI stain, you will probably need to manually adjust the smoothing filter size and the maxima separation distance. The default smoothing filter size (~6 pixels) doesn’t look too bad, but the maxima suppression distance should probably be larger; a value of 10 pixels seems to do a better. Of course, theses suggestions are subject to your further tweaking.

Also, it seems that a portion of the image is somewhat saturated; this may be due to detachment of the cell layer at an edge, an will affect nuclei identification in that region. If it is not biological in origin, I suggest checking your image acquisition settings.

The second item I would recommend is that for the CD45 and DsRed images, for the purposes of identifying positive staining, you only need to properly threshold the image and not worry about segmentation (i.e., declumping); you can disable the size and image edge exclusion, as well as setting the declumping to “None” and disabling hole filling.

Regards,
-Mark