Count number of GFAP+ cells (costained with DAPI)


I’d like to be able to count the number of GFAP+ cells (red) that I obtained in my mixed culture after having costained using DAPI. I’ve attempted to use the example pipeline (ExamplePercentPositive) but have been unsuccessful. Can you provide me with details on how to go about doing this? I’ve changed parameters in NamesAndTypes to match my files but it still doesn’t work and I keep getting the following error message: “the pipeline did not identify any image sets.” I did not however press “Start Test Mode” which I’ve read in the forum might cause this. I’ve been inputting two files (separate DAPI and GFAP channels) which doesn’t seem to work. I’ve also converted to 8-bit in Image J prior (first for just the DAPI, and then tried again converting both to 8-bit).
Thanks for your help!



Make sure the file names of your images contain information on which fluorescence channel they were obtained in (e.g. “001_dapi.tif, 001_gfp.tif, 002_dapi.tif, 002_gfp.tif”, etc. You could also place them in seperate folders (one “dapi” and one “gfp” folder).

Select the files/folders, drag&drop them into your “Images” module.

The NamesAndTypes module needs information on how to identify different channels. So if you added “dapi” and “gfp” to your filenames, you could try this

Press the update button below (its not visible in the image), and see what happens. If you get an error, make sure all above is true.

That should solve your error message (and if you get an “Zero Length Selections not Allowed” error, it will still work).

However, the rest of the pipeline would depend on your images. As GFAP is extranuclear (?), you will need to identify the cell bodies to determine if the cell is positive or not. I think you could try the “Human cytoplasm-nucleus assay (SBS Vitra)” example pipeline to measure the cell body intensities. Alternatively you could identify the nuclei and expand them a little with the “ExpandOrShrinkObjects” module, and measure the resulting object.

Trial and error, don’t be discouraged!


Thanks @Fabba123 . And yes @glia, can you please post some sample images and your pipeline here?


@Fabba123 I had tried altering the file names so that they matched what was in NamesAndTypes to no avail but your advice to keep it simple (“dapi” & “gfp”) is what ultimately ended up working. Many thanks for your help!
@David_Logan I kept the pipeline pretty much the same and only tweaked the thresholding parameters after auditing the output to have CP correctly identify the cells. @David_Logan On an unrelated note I’m unable to open up CP on my iMac that has OS X El Capitan. It does however work on my OS X Yosemite. I see others using El Capitan have had issues but I’m not able to really determine how to get around this from those threads. Sorry for inserting another question here.


@glia You can try the Beta here, which should work with El Capitan: