Speckle Counting pipeline

Dear all,
i’m trying to quantify the numbers of foci obtained by h2ax technique. As you suggested i’m trying to view initially the examples and then going to evaluate my case study. however in your page cellprofiler.org/examples.shtml , the images that appears in the example had no cells, are all black so the after runing the example I don’t have any results.
however i tried to use my own pictures and the results were the same, so i maybe the problem is mine :smile:
To perform this pipeline as far as i could understood i need two gray images. IOne taken with a dapi filter giving us the nucleous and other one with FITC for example giving us the highlithed spots, isn’t it? when I run bith picture using the pipeline of the example, the final results, i.e. the graphics don’t give me any information, it appears all blue without any curve, spots, lines… Could you hel on this.

Many thanks in advance. Best regards,

Ana Belchior, Lisbon, Portugal

Hi Ana,

The example pipelines are intended to act as a starting point, so while they may be similar to your assay, it is typical for the settings to require additional adjustment for your specific images.

That said, it seems that thresholding settings need to be tweaked so the desired features can be detected correctly. If you like, could you post a couple of sample images so we could take a look?

Regards,
-Mark

Hi Mark,

Thanks for your fast answer. Herewith, I send you 3 images. DAPI+FITC indicates an image auired with both DAPI and FITC filters, FITC it’s an image acquired only with FITC filter and the tirth one was acquired with a DAPI filter.

If you need any more details please let me know.

Thanks one more time.

Ana






Hi Ana,

Attached is a pipeline which should work, and is not very different from the example speckle couting pipeline.

However, I should note that your DAPI image has a severe illumination problem, leading to a large intensity variation across the image. I’ve attempted to correct for this using the CorrectIlluminationCalculate and CorrectIlluminationApply modules, but if this is indeed a typical image, I strongly suggest checking or improving your image acquisition protocol. In addition, your foci are not very distinct, causing the EnhanceOrSuppressFeatures filtering step to fail to enhance the speckles as well as it ordinarily would. Is there any way to improve the image at either the image acquisition step or the staining step?

Regards,
-Mark
2011_02_22.cp (8.84 KB)

Hi Mark,

Thanks one more time. I understand what you are saying, my images aquisition is not good I have to improve it. So, as far as I understood your opinion, my DAPI is to shining and on contrary my FITC cannot show distinct foci.

regards,

Ana

Yes, that’s pretty much correct. As a counterexample, refer to the images for the example speckles pipeline:

  • For the nuclei image, there is good distinction between foreground (i.e., nuclei) and the background across the image, making nuclei detection fairly easy.
  • For the h2ax image, the foci are distinct and clearly punctate. Since I don’t know your phenotype, the haziness of your foci may be a consequence of the biology, but sometimes optimizing the staining protocol plus the focus can help things.

Regards,
-Mark

Hi,

Yes, I understood the way the specklespipeline identify the foci, basically we need to give to the program two images, one with the nuclei and other one with foci for a correct localization of the foci and of course for the program coukld attribute each number of foci to a specific nuclei. Then, other question comes out: When i run the pipeline i have my results in a excel file, until now everything is ok :smile: the output tell me the total number of nuclei, and then the total number of speckles, and what about unique information for each nuclei? I have to use the .mat file? Is there a way to look at this file?

All the best from Portugal,

Ana

In the ExportToSpreadsheet module, if you click the “Add another data set”, you can add the nuclei for export. Keep in mind that without a Measure module to measure something about the nuclei, all the spreadsheet will contain about the nuclei are their labels and locations. You can always insert a MeasureObjectIntensity or MeasureObjectSizeShape module to record more information about the nuclei.

Regards,
-Mark

[quote=“mbray”]In the ExportToSpreadsheet module, if you click the “Add another data set”, you can add the nuclei for export. Keep in mind that without a Measure module to measure something about the nuclei, all the spreadsheet will contain about the nuclei are their labels and locations. You can always insert a MeasureObjectIntensity or MeasureObjectSizeShape module to record more information about the nuclei.
[/quote]

Hi…

I am trying to analyse the h2ax foci in cells,
Can you please explain in detail how to get the number of foci/nuclei? Also can I delete the modules which measure intensity of Nucleus and H2AX? since we only need to count the foci and not measure the intensity (i dont know if intensity is necessary for the calculations)

Thank you very much for an excellent software…

Pavan

Hi,

RelateObjects stores the counts of child objects (foci here) in the Parent’s table. So look in the ‘Nuclei.csv’ in the ExampleSpecklesImages output folder for the “Children_h2ax_Count” column.

Yes, you can delete the measurement modules. RelateObjects only needs the location of the objects which is saved by the Identify* modules.

Good luck!
David