Wave 3 Channel Compensation Results Error

[jenna-tomkinson]

When testing with the CompensateColors module in CellProfiler to perform channel/signal compensation for the Mitochondria and Gasdermin channels (which are most likely to have overlap between each other), the results are strange and unexpected.

The figure below shows on the top row how the Mito and Gasdermin channel images are before compensation (it is very obvious the bleedthrough in the Gasdermin channel) and the bottom row shows the results after compensation (MitoComp has been artificially brightened as to be able to view the specks) .

In the MitoComp image, there are small specks across the image which are expected of the Gasdermin channel. While in the GasderminComp, there are structures that are expected in the Mito channel.

I have confirmed in an issue with the CellProfiler team that the module parameters must be set where each channel has its own class. This leads me to believe that something is happening within the image acquisition stage.

For the time being, it is best to not include the channel compensation module with the analysis pipeline until this is resolved.

[gwaybio]

For the time being, it is best to not include the channel compensation module with the analysis pipeline until this is resolved.

Thanks for troubleshooting this @jenna-tomkinson ! I think you should reach out to Milad - It will be good to confirm channels... I suppose there is also a tiny chance that the compensatecolors plugin is swapping images! 😬

Could you also show an artificially brightened MitoComp? It'll be good to visualize the specs a bit more clearly.

[jenna-tomkinson]

@gwaybio The image has been brightened and I started drafting the email right after making the issue 😄 I will be sending it in a few minutes. What I can do is try this module with the other channels and see if there is any channel swapping in the meantime.