Select the cells with cytosol staining marker

cellprofiler
identifyprimary

#1

Hi all,

I am trying to generate a pipeline that select the cells with high cytosolic staining specifically and then measure those cells’ nucleus and cytosolic signal intensity.

Since not all the cells have the cytosolic marker or not even signal intensity, I have tried to “Threshold” the signal followed by IdentifyPimaryObjectes with the Hole filled option checked in hoping to identify the cytosolic marker positive cells then we could identifysecondaryobject. Somehow my pipeline could not capture the cytosolic marker positive cells easily…

Here are several representative pics: https://www.dropbox.com/sh/gobhaxe1wn1wb3w/AAB4UPoDuqHJarx8o8c3LS82a?dl=0
in which green is the cytosolic marker, blue is dapi, and red is another nucleus staining.

I am wondering if my algorithm design is correct or not or is there any better way to do this?

Thanks, FY


#2

Hi there,

I have tried to search the forum and found some useful info.

Now I could have the cytosol positive cells selected, but I am still finding a way to use the selection from cytosol to back select the corresponding nucleus.

Here is the pipeline and project file attached. Also have the images with the dropbox link…

Thanks, FYtesting1.cppipe (12.6 KB)


#3

Hi,

I think what you were missing the existence of is the FilterObjects module. By filtering only cytoplasms that grew out a certain amount in your IdentifySecondaryObjects step, you can select only nuclei from positive cells, which you can then use to more accurately ID the cell boundaries afterward if you like. Try it out and see!

testing1_BCEdits.cppipe (15.4 KB)


#4

Hi Beth,

Thanks so much! It really works pretty good!

Just one thing I am not so sure here:

in the FilterObjects step, we are filtering out the area value less than 100 as shown in the snapshot, just wondering how would we determine this value?


#5

I did it empirically- feel free to change it to something else if you want to be stricter or less strict on which cells to count as “positive”.


#6

Thank you so much!

PS: Seems you are on fire today! I have seen millions of replies from you!