How could I detect longitudinal myotubes


I am new to CellProfiler and got stuck with my analyzes.

Background: I have Myotubes on a cellular level which are expressing eGFP and are stained against alpha actinin and DAPI. I would like to compare the levels and localization of GFP expression to alpha-actinin. In my opinion DAPI won’t help that much, because the myotubes are multinucleated.
So far, I could manage to identify alpha-actinin as my primaryobject but I am stuck with my secondary object GFP. I am able to detect most GFP expression, nevertheless the strongest expression is not detected – somehow feels like it is above a certain threshold. I already played around with lots of settings. Still I was not able to detect the brightest GFP expression. All topics related to muscle-staining have a cross section and no longitudinal tubes.

Thanks in advance for your help, I will support my pictures in the second post.

Related pictures:

Cy5.tif (8.0 MB) FITC.tif (8.0 MB)

Hi @Mahuesi,

I’d be happy to take a look at your pipeline to see if I can find a few improvements. Could you send your pipeline file?

Great work so far!



thanks alot for your answer. Attached you can find my pipeline.

Bestmyotube analyze.cpproj (658.7 KB)

From the example images and pipeline that you sent, I don’t see the error that you’re discussing above (bright GFP + cells are not detected by IdentifySecondaryObjects). I think I can see where the error might arise in your, general strategy, though.

Your current pipeline creates objects based on the alpha-actinin channel in IdentifyPrimaryObjects. These objects are then used as seeds to grow new objects where the new edges are based on the GFP channel in IdentifySecondaryObjects. If a given cell that is strongly GFP+ is not identified in IdentifyPrimaryObjects based on the alpha-actinin channel, then the cell can’t be identified with IdentifySecondaryObjects (since no seed exists to create the object). One solution to this challenge is to adjust the thresholding on IdentifyPrimaryObjects in order to ensure that seeds are created for every GFP+ cell that you’re interested in. Alternatively, if your GFP stains every cell, it may be possible to detect the cells directly using an IdentifyPrimaryObjects module.

These concepts are explained more fully in this CellProfiler introductory workshop, which is available on the Center for Open Bioimage Analysis YouTube Channel. You can download the corresponding written tutorial on Translocation from the CellProfiler Github page.

Hope this helps and let us know if you have more questions!

Thanks alot for your help!

CellProfiler is such a helpfull tool.

All the best and stay healthy.

Hi everyone,
finally I solved the problem: I use two times IdentifyprimaryObject and identify with this GFP as well as actinin. After that its easy to compare the values.

Thanks for all your help!