How does CP determine "Nuclei_Granularity_5_Mito float"?


#1

Hello CPForum members,

I would like to know how CellProfiler (CP) measures "Nuclei_Granularity_5_Mito float"
I am doing Cell Paint assay with attached pipeline and using CP (My CP version is 2.2.0).
final_analysis.cppipe (41.4 KB)

I read this post and I understand the feature value called “Granularity” in CP is the distribution of granularity and the number like “_5” mean the size of object.

Then, I think “Nuclei_Granularity_5_Mito float” means the granularity distribution of “Mitochondria” whose size is “5” in the region of nucleus (the region of nucleus is determined with “IdentifyPrimaryObjects” module).

I don’t understand the meaning of mitochondria in a nucleus from the view of biology because there is no mitochondria in a nucleus, but is my understanding for the feature value called " Nuclei_Granularity_5_Mito float" right?

Best wishes,


#2

Hi,

In this case, “Mito” is referring to the channel, but your overall unpacking process of what the different things mean is correct:

Nuclei means were’re measuring something about the nuclei

Granularity means we’re measuring the nucleus’s granularity

5 means we’re looking at granularity at the scale of 5 pixels (as opposed to 2 pixels, or 20 pixels)

Mito means we’re looking at the nucleus’s granularity in the Mito channel, AKA the Alexa647 channel. You’re right that given what we know from biology there’s unlikely to be a ton of intensity at all in that channel in the nucleus’s location, but a) there might be mitochondria that are over or under the nuclei that DO get picked up and b) for cell painting we always err on the side of measuring MORE things so that we are more likely to pick up unusual or subtle phenotypes.

Does that help at all?


#3

Hi bcimini,

Thank you for your lightning quick response.
Your answer is very helpful for me to understand this topic.
For me, CP is excellent software which provides a lot of hints about phenotypes and it is
the biologist’s task that translate those hints to biological meanings.

Then, could you let me check one more thing?
How does CP measure the nucleus’s granularity in the Mito channel??

I can find nucleus in the images taken in Mito channel by my eye because there is contrast between
inside of nucleus (dark) and outside of nucleus (bright). How about CP?
I would like to know how CP looks at the nucleus’s granularity in the Mito channel.
Understanding principles how CP calculates feature values allows me to interpret the results from CP properly.

Thanks,


#4

I can help! You identify the nucleus using the DNA channel, creating an outline around each one. And because the DNA channel and mitochondria channel line up with each other, CellProfiler “transfers” the outline of the nucleus from the DNA channel onto the mitochondria channel, and then goes ahead and measures whatever mitochondria signal it can find in the nucleus region.

Does that help?


#5

Hello Anne,

Thank you for your reply and explanation.
That is what I want to understand!

I will link the feature values with biological meanings.

Best,