Nuclear mask excluding nucleoli

nuclei intensity excluding (492 KB)

I want to quantify the nucleoplasmic signal of my protein of interest but excluding the staining in the nucleoli (structures inside the nucleus). I have experience in generating pipelines to identify objects inside the nucleus and measure properties and relate it to the parental nuclei. However I have never used those objects inside the nucleus to create a “negative mask”, meaning a nuclear mask that excludes them. I have sets of 3 pictures: channel 1=DAPI, channel 2= my protein of interest, channel 3= a protein that stains nucleoli in order to identify these structures. Is it possible to generate this kind of nuclear mask without the nucleoli? Or is it better to quantify the signal in the whole nucleus but then subtract the signal from the nucleoli? if this is the case, how can I subtract the signal of all the nucleoli to the same parental nucleus? Also, I would like to keep the data per individual nucleus.
Please, can you help me?

Thank you


Hi Barbara,

I think any of your approaches might work fine. My choice (attached) is to identify and mask the nucleoli, then identify the protein outside the nucleoli. You could try and subtract the images (protein minus nucleoli) but I think this method outlined below gives you more fine control over the various thresholding and segmentation steps.

So the steps are:
(1) Gray To Color: just for visualization - you can disable this module by clicking the checkmark.
(2) ApplyThreshold: Creates the mask of the nucleoli
(3) MaskImage: Applies the nucleoli mask to the protein image
(4) IDPrimary #1: Segment the protein outside the nucleoli
(5) IDPrimary #2: Segment the nuclei
(6) RelateObjects: Relate the protein found to their parent nucleus.

In the Parent object CSV ("…Nuclei.csv"), you should find the count of the protein spots outside the nucleoli, per nucleus. You’ll have to quantify whatever else you want.

Hope that helps!
DL_exclude_nucleoli.cppipe (12.5 KB)

1 Like