Measuring Object Intensity


#1

Hello,

I am quite new to cell profiler. I work with monocyte derived macrophages and I’m trying to compare the expression of pro-inflammatory and anti-inflammatory markers produced by the cells. I stained the cells for the nucleus (DAPI_blue), IL-10 (FITC_green) and TNF-alpha (Rhodamine_red) which are both in the cytoplasm. I created a pipeline to measure object intensity. I use the nucleus as my primary object to located the cells; I have two secondary objects, Cells for TNF signal and Cells 2 for IL-10 signal. However after running my anaylsis, I realized both Cells and Cells 2 measured the green and red object intensities.
I expected Cells to measure the red intensity only while Cells 2 measures green intensity only… Did I do something wrong?
This is my piplepline
trial pipeline 2_integrated.cppipe (24.0 KB)

This is an image set

Thanks for your help.
Didi.


#2

Hello there,

In your current module MeasureObjectIntensity you let dual objects pair with dual channels, it will result in 4 total measurements.

If you expect “Cells to measure the red intensity only while Cells 2 measures green intensity only”, then it should be set in 2 separated modules:

Otherwise, your pipeline looks great.

Bests,
Minh


#3

I get it now!

Thanks so much Minh.

Best wishes,
Didi.


#4

Hi there,
I’m new with CellProfiler. I have 2 channel image: Dapi and a nuclear marker in 488. I need to measure the intensity of the 488 positive object, but at the same time I need to have the value of dapi intensity for each identified object. Is it possibile? I tried with different approaches, but I not had good results. In fact I obtained an intensity of the all dapi objects, that are different compare to the intensity value of dapi object identified in 488 positive cells.


#5

Hi,

Let’s try to work it out together:

  • Let’s first identify the nuclei by IdentifyPrimaryObjects, using DAPI channel, let’s call them objects “Nu”, for e.g. we now have 1000 objects “Nu”.
  • Then we measure the DAPI signal (MeasureObjectIntensity) in every Nu, now we have 1000 values.
  • Then we measure the 488 signal (MeasureObjectIntensity) in every Nu, we still have 1000 values, but about 700 of these values are high for 488-positive Nu, and 300 of these values are low for 488-negative Nu.

This way, the values of DAPI signals are independent from being 488-positive or 488-negative.

(Do NOT identify your objects by using 488 signal)

Hope that helps.


#6

Thanks Minh.
I tried as your suggestions and I obtained this excel data:


In same case, I obtained very high values (as you can see in the highlighted line): how can interpret them? At the end of this analysis I want to represent every single object in a graph with a scale from 0 (negative 488 object) to 1 (100% 488 positive) , where the 0 object are the negative for 488, 0.5 an object with a average positivity, etc…so I need also the 488_positive intermediate values.
I need also some suggestions to put a threshold to define the 488-negative objects ( I obtained only 2 objects that are 0.0, but a lot of them had a 0.000XXX value, in fact the 488 positive objects are 101 on 113 total Dapi objects).
Foxg1_488.cpproj (863.6 KB)

Thanks for your helpful advices.
Angela


#7

Hi,
In Excel, you can try to select the 2 columns Intensity_MeanIntensity_DAPI and Intensity_MeanIntensity_488, then make a scatter plot out of them. You will then see a distribution of 488-positive/negative to decide where is the threshold.

Hope that helps.


#8

The intensity should always be a number between 0-1; I’m wondering if somehow the values got corrupted opening the CSV and that should have been 7.40 x 10^-11? Can you look at that nucleus and tell if it’s pretty dim in that channel?


#9

Thanks for your help