Would need some basic help


Hi everyone,

I´m completely new to the program and I have some problems to build the pipeline and especially to define the parameters I would need to evaluate my pictures.
I´m not even sure, if I can really evaluate my cells with the program, maybe someone could help me here with some advice.
What I want to do is to count autophagosomes in my mouse embryonal fibrolasts which express a GFP-LC3 fusion protein. I attached an example how these cells look like. One of the problems might also be the diversity in the size of the autophagosomes.
I tried to change the example “Speckle Counting” and to adjust it to my purpose but wasn´t successful with this.

If anyone would have a hint or advice how to do this, this would be great.

Thanks :smile:



Thanks for giving CellProfiler a try! Based on your images, it should be more than adequate for your needs. Try a pipeline with the following modules:
(1) IdentifyPrimAutomatic to identify the Nuclei
(2) IdentifySecondary using the Nuclei as the primary objects and the Cells as the secondary objects (I think Propagation as the identification method and Otsu or MoG as the thresholding method should work)
(3) CorrectIllumination_Calculate, using the same values from the same module in the ExampleSpeckles pipeline (your images seem to be about the same size so it should be OK)
(4) CorrectIllumination_Apply upon the original autophagosome image and the illumination function previously defined
(5) Crop the corrected image into the Cell shape (select “Other” for the third box and type the name of the object you gave the Cells)
(6) IdentifyPrimAutomatic on the cropped corrected image to identify the Autophagosomes using the per-object thresholding method of your choice (MoG per-object seems to work well)
(7) Relate the Autophagosome objects to the Cell objects
(8) ExportToExcel to get your data out

As you can see, what I’ve described is not that much different than the Example Speckles pipeline other than the use of the secondary objects. The part you might need to adjust the most is either the parameters in (4) to get the best contrast of the autophagosomes or the thresholding method in (6) to segment the autophagosomes correctly.

I think this should get you started. Hope this helps!


Hi Mark,

Interesting that you’d put the CorrectIllumination modules after identifying nuclei and cell boundary.
In my pipeline, I’ve placed this before any identify modules.

What would be your reason in this case for doing this?

Cheers, Tim


Hi Tim,

This is a case in which I’m co-opting the functionality of CorrectIllumination_Calc/Apply for something it was not technically intended for, i.e., highlighting the auotphagosomes rather than correcting an image for optical aberrations.

My intent here is to smooth the image out just a bit and then subtract it out, relying on the fact that the autophagosomes have a sharper intensity gradient than the surrounding cytoplasm. Subtracting the smoothed image from the original would remove much of the “background” (i.e., cytoplasm) while leaving much of the autophagosomes intact enough for detection via thresholding.

Possibly, SmoothOrEnhance using tophat filtering could also do the job, but I don’t think it would do it as well.



Thank you for your help :smile:
It already works quite well if I just run it as you proposed and I think it will do exactly what I need, after I did some little adjustments.

Thanks a lot!