Beginner Colocalization Pipeline Adaptations and Questions




My name is Alex and I’m new to using CellProfiler. Briefly, I’m a neuroscientist who does Fluorescent In Situ Hybridization and am looking for a good way to quantify our data using CellProfiler. It seems to do a great job with what we need, although I am struggling with the settings for some of the pipeline and I’m hoping I could get some help. This is all taken straight from the colocalization example pipeline provided online and parameters adjusted slightly.

I’m including some pictures with this post to show my dilemma. First off, we have two channels we’re looking at: CTB-555 (Soma) and DAPI (Nuclei). One picture is the whole image view (high-res Grey-Scale MAX projection Z-stack TIFF taken on a confocal), one is a zoomed in view of our area of interest, and one is of the settings I’ve used in the IdentifyPrimaryObjects module. (NOTE: I could only upload 5 pictures, so the settings for the DAPI were the same as in the example pipeline except typical diameter range changed to 5-10)

If I’m not mistaken, this module is an especially critical module to get right since that really sets the detection settings that will determine what gets selected in each channel before processing and quantification. Based on the few detection settings I changed it definitely got better in identifying the Soma in the CTB channels, but noticed it still missed some cells. If you look at the zoomed photo (Soma_Zoom) you can see cells near the middle of the image that are outlined in purple that aren’t counted as CTB+.

My question is what settings should I focus on changing to better outline/identify this image? Changing the diameter settings seemed to help, as well as the threshold, but since our computers take a few minutes to run the image processing, I was hoping to avoid the process of trial and error and rerunning the analysis to see how the IdentifyPrimaryObjects settings changed the output. If there are other processes upstream up this that also could use adjusting, let me know and I’ll send over those settings too.

Any suggestions for an efficient way to optimize settings would be greatly appreciated! Even just which parameters I should look at or change would be very useful. In the end, I will be trying to colocalize 2 other channels in addition to the DAPI and CTB, but I thought I’d start off with something a bit easier.

Thank you all for you help!



You can play around with the tunning of Smoothing filter size and suppressing local maximum options for declumpling and merging of objects. Objects in purple color are not detected as they are off to specified diameter range. Click on image and you can find a meaure length option when you scroll down tools at top navigation bar. This will help you to determine range of objects.



Thank you Hamdah! I will definitely give some of these setting options a try and see if it gets better.

Many thanks for all your help!