GFP fluorescence intensity in Cy5-identified Objects


I am using CellProfiler to characterize bacteria containing vacuoles in infected epithelial cells. I have GFP-tagged bacteria and I am using a Cy5-tagged antibody (Lamp1) to identify the bacteria containing vacuoles. In addition to that I have Hoechst to label host cell nuclei and I am using a Cy3-tagged celltracker dye to label my epithelial cells.
I have set up a pipeline to identify (plenty of) objects in my experiment and I can clearly identify my Cy5-labeled bacteria-containing vacuoles as well as GFP-labeled bacteria.
I’m exporting all data in an SQL lite database so that I can then use the Classifier in Analyst to rank “normal” versus “small” or “large” vacuoles etc…so far so good.
Now, what I would like to do is to calculate the GFP intensity WITHIN my Cy5-labeled identified objects as a way to determine “how many” bacteria I have in each vacuole. I’d like to have this measure included in the characteristics of my bacteria-containing vacuoles so that it would be a parameter in my “vacuoles database” and Analyst can use it to further classify the vacuoles.

Any idea on how to do so? Is there a way to tweak the MeasureObjectIntensity module so that I can calculate the GFP intensity of previously identified Cy5 objects?
Or alternatively (it may be easier…) to use two databases (the one containing the infos on my vacuoles and the one containing the infos on my bacteria) in Analyst Classifier?
If it can help I am attaching my pipeline

Thanks in advance

siRNA Screen Higher Threshold Colonies.cpproj (1.5 MB)

Hi Matteo,

If I understand what you want correctly, there are three approaches you can take:

  • Measure the per-vacuole GFP intensity overall: Use the MeasureObjectIntensity module with the vacuoles as the input objects and the GFP image as the input image.
  • Measure the per-vacuole GFP intensity from the bacteria: Use MaskImage to mask the GFP image with the bacteria objects. Then use MeasureObjectIntensity with the masked image as input so that only the pixels from the bacteria will be considered.
  • Measure the mean bacterial intensity on a per-vacuole basis: Use MeasureObjectIntensity with the bacteria as the input objects and the GFP image as the input image to obtain the per-bacteria intensity measurements. Then use RelateObjects to match up the bacteria to the enclosing vacuole and select “Yes” for the “Calculate per-parent means…” setting. This will take the average of all the bacterial measurements on a per-vacuole basis.

All three of these approaches will assign the measurements to the vacuole objects. What’s nice is that you can even use all three of these together to have an extensive set of measurements for the machine-learning tool in CellProfiler Analyst to work with.



Looks like it’s working perfectly!