Nucleus and cytoplasm outlining in yeast cells

Hi,

I am new at using CellProfiler and I am trying to quantify green and red fluorescence in the nucleus and cytoplasm of yeast cells. Having read one of the questions and answers in forum, I managed to do a decent outlining of the nucleus and this is working fine I think. However, the outlining of the outer cellular membrane is still a bit problematic, the outlined cell tends to be larger than the cell itself. I am attaching two of my images (green and red channels) and my pipeline. Could you please tell me how I could improve it, so I have a better delimitation of the cell?

Thanks a lot!

Michele




Yeast cells_Nucleus and Cytoplasm.cp (22 KB)

Hi Michele,

I would use IdentifySecondaryObjects with the green channel, with the nuclei as the input object (which looks good, BTW; nice work). Since I’m not sure what you deem as too big or too small (since I’m not the expert on the biology), I’d recommend starting with Otsu Global with 2 classes. If that doesn’t look good, then switch to 3 classes with the middle class set to the foreground if you want to capture more (i.e,larger cells), or set to the background if you want to capture less (smaller cells).

Regardless of whether you are using two or three classes, you can always use the threshold correction factor to adjust the threshold upwards (by increasing the correction value) or downwards (by decreasing it) if you find that a particular choice of thresholding method is consistently too stringent or too lenient.

Regards,
-Mark

Hi Mark,

Great, thanks a lot for your reply. I’ll play around with your suggestions and see how it goes.

Also, great software! Its really powerful once you start getting the hang of it.

Best,
Michele