Cytoplasm-nucleus ratio


#1

Hi all!
I am new to the CellProfiler software. I tried to run the SBS pipeline in my computer. What I got were 5 excel files. Do we need to do some calculation by ourselves to get the cytoplasm-nucleus ratio? Or could we find the ratio directly in the spreadsheets?
Looking forward to your reply,
Thanks in advance!
JJ


#2

Hi,

Great question! Congrats on getting that whole pipeline run, first of all- its not short. There are two ratios automatically generated by the SBS pipeline- Ratio1 and Ratio2 (these are created by the ‘CalculateMath’ modules near the end of the pipeline). If you open either of the DefaultOUT.mat_Nucleus.xls or DefaultOUT.mat_DistCytoplasm.xls, these two ratios will be the last two columns (so you have to scroll waaay to the right). Since these measurements are a ratio of nucleus and cytoplasm measurements, they’re stored in both Excel files (otherwise, just Nucleus measurements are in the Nucleus excel file, etc).

Ratio1, as defined by the first CalculateMath module is the ratio of total Nuclear intensity/total cytoplasm intensity. Ratio2 is the Mean Cytoplasm intensity/total nuclear intensity. You can change these ratios to whatever measurements you want- Mean nuclear intensity/mean cytoplasm intensity, etc by changing the settings in CalculateMath.

Hope that helps!

-Kate


#3

Dear Kate,
Thanks very much for your prompt reply! I really appreciate that.
I found those ratios in my spreadsheets. It seems that the programme automatically genenrates those 2 ratios for each cell in a image. Is it possible that we can calculate a mean nucleus-cytoplasm ratio from hundreds of cells measured in one single image?
I have another question. In your reply, you said that ratio2 is the Mean Cytoplasm intensity/total nuclear intensity. I just don’t understand how the mean cytoplasm intenisity or mean nuclear intensity is measured, since the measurement is for one cell only.
Finally, I would like to ask, is there any way to speed up the running process? It gets slower after running for a few cycles. So I needed to seperate the example images into different folders and run the software for several times. I have around 500 photos for each experiment. I think it will take days to analyze those images.

with best regards,
JJ


#4

Hi JJ,

To get the mean value of the ratio for an image, the ‘Image’ Excel file will contain those data. Again, you’ll have to scroll way to the right (in my Excel file, it was columns DT and DU- or you can just search for ‘ratio’). This is the mean value of all the ratios for an entire image.

For each cell, the mean and total intensities are defined based on the segmented cytoplasm or nucleus from IDPrimAutomatic, IDsecondary and IDTertiary. Total intensity (or ‘integrated’ intensity) within the nucleus is the sum of all the pixel values contained within the region defined as the nucleus from IDPrimAutomatic; mean intensity is simply total intensity divided by area. Depending on the stains you use or your particular assay, you may find one measure is better than the other for describing translocation.

For tips on how to speed up processing, in CellProfiler check out Help>General Help>Memory and Speed and see if any of those suggestions help.

-Kate