Why the cytoplasm-to-cell ratio is bigger than one?

Hello,
I am trying to use CellProfiler 2.1.1 to calculate the nuclei-to-cell and cytoplasm-to-cell ratio of the mean intensity of a protein.

Thank the team a lot for the great software.

My question is why the cytoplasm-to-cell ratio of the mean protein intensity was bigger than one? And why the sum of the ratio of cytoplasm-to-cell and nuclei-to-cell was not equal to one?

Thanks

Hello,

The mean intensity tells you about the average concentration of your protein in the specific compartment where you measure it; this can be higher in the cytoplasm than in the whole cell, e.g. if your protein resides more in the cytoplasm than in the nucleus. Thus mean_cyto/mean_cell can have any value >= 0.

The integrated intensity however tells you about the total amount of the protein in the specific compartment. Thus integrated_cyto/integrated_cell is a number between 0 and 1. And then indeed:
integrated_cyto/integrated_cell + integrated_nucleus/integrated_cell = (integrated_cyto+integrated_nucleus)/integrated_cell = 1

In my experience for most biological questions the integrated intensity is the better readout than the mean intensity.

1 Like

Thank you, Christian.

I see now.

I changed the setting of MeanIntensity to IntegratedIntensity in the Module of CalculateMath, the ratio value of integrated cyto/integrated_cell (0.49) or integrated nuclei/integrated cell (0.57) was between 0 and 1 now. However, the value of the integrated_cyto/integrated_cell + integrated_nucleus/integrated_cell was still not equal to 1.

Do you have any idea about this? If possible, could you please help to check whether there is something wrong with my settings?

Many thanks

Qinchang Zhu

0504 Sample (integrated).cpproj (847.5 KB)

Hi,

  • In IdentifyTertiaryObjects you checked “Shrink small objects prior to subtraction”; I guess this could result in the total area of nucleus and cytoplasm being larger than the area of the cell, so the ratios would not add up to 1. So, you could try what happens if you uncheck this option. Using CalculateMath and MeasureObjectSizeShape you could also directly check that the areas add up properly.

  • When checking whether the ratios add up to 1 you have to check this for individual cells, e.g. in the exported “cells.csv” table; if you check this for mean values of cells in an image, e.g. in the “image.csv” table, this does not necessarily add up to 1.

  • In general to figure out what is going on, I recommend checking the per cell values using DisplayDataOnImage in CellProfiler.

Does this help?

Hi, Christian,

After unchecking the “Shrink small objects prior to subtraction” in the Identify TertiaryObjects, the sum of ratios from N-to-Cell and Cyto-to-Cell became 1, no matter for the mean ratio for all cells or ratio for single cell.

Thank you so much.

1 Like