Help with Nuclear translocation pipeline?

cellprofiler

#1

Hello everyone, I would like to say thanks up front!

I have a pipeline I have made and some images, but as a new user I can not upload it :frowning:

  1. I would like to setup a way in the pipeline to determine the average intensity of the uninfected samples, then use this + 1 standard deviation to set a threshold cutoff for what i consider p65 translocation. Then apply this to the cells and generate a % translocated above uninfected calculation as a separate output file.

  2. I would like to have a crop, background subtraction, illumination correction, and a merge for my channels as individual channel images and the merged. I would also like to save the outline of the nucleus and the cell on top of the gfp channel.

  3. I would like to calculate in addition to just median intensity the % translocation from the cytoplasm to nucleus. Then generate a image overlaid with this “localization” / “threshold” that reflects intensity?

  4. Help refining the de-clumping/ segmentation of the nucleus - or a way to discard/ or identify ones that are not individual cells via a label on the overlaid image that corresponds the intensity output file

At anyrate, I would love to have some help and feedback! This is not my first time using cell profiler, but is the first time i am trying to build an extensive pipeline.


#2

Hi,

Have you checked out our Translocation tutorial? It has many of the things you’re asking for.

As for 1), you can’t have CellProfiler run all images, make a calculation, then run all images AGAIN; you’ll need to run one pipeline to do the calculation you want, and another to apply it.

As for the others, I think if you search the issues individually you’ll find guidance on how to do them (search the program help, too, for words like “Overlay”); you can always upload your pipeline to GDrive/Dropbox/etc and link it here.

Good luck!


#3

Hello, thank you for your response. I have been working on the pipline with your suggestions. The dropbox link for the pipeline and some test images is below. Any further suggestions would be great!

I am currently struggling with understanding why my intergrated intesnity levels for the positive control is so low. When I use FIJI to threshold/watershed for the nuclei and apply these as ROIs to the p65 image I get more believeable data (IE, the intergrated intensity for the positive control over all the nuclei is always 2-4 times higher than the negative control). Whereas, when i use the intergrated intesity values from cell profiler there is only 20-50 unit difference in the mean.

A6 and A11 are positive control for nuclear translocation
A7 and A12 are negative control


#4

Hi,

Without knowing how you do your FIJI analysis, I can’t say for sure why or how you’re getting the results that you are, but I’m suspicious that you see a 2x increase in translocation between the image on the left and the image on the right- neither looks particularly translocated to me, certainly not 2X.

Your images aren’t all of the same mean brightness either- just looking briefly at the outputs of MeasureImageIntensity, the ones in well A7 seem to be about 20-40% dimmer overall than the ones in well A6.

Finally, unless you expect nuclear size to be an interesting change to be measured by your perturbation, I would use mean or median rather than integrated intensity, as the latter contains a measure of size. Even better would be the ratio of mean intensity in the nucleus to mean in the cytoplasm.