How to calculate transfection efficiency



I am a new user and I just started to learn CellProfiler. I took some cell images. There are one or two cells per image which are successfully transfected with DNA fused with GFP. And the GFP protein forms accumulation in the cell. I did not stain the nuclei, but I wanted to calculate transfection efficiency using CellProfiler. Could you please tell me which modules in order should I set?

I have another question: I first loaded the images, and then run the color to gray module. In this module, I used combain and only the green channel. But when I want to run the next module, it is said that it cannot find the the images which is already turned to gray. I also tried the split in the module, but I always get this problem. Do u know what is wrong?

Thanks very much in advance.


May be you can do primary automatic to identify cells and crop this cell area from second image and try to identify signal depending on background setting… That will make job easier. Once you have cells identified and you know how many are positive as per your settings, you will be able to decide transfection efficiency… But Cell profiler team may suggest better way to do it…


Hi there,

The pictures you uploaded seem to be brightfield images merged with a fluorescence image. Is this the case? If so, you are probably better off starting with the images as two separate channels rather than merged, because the first step of your pipeline will have to separate the colors. If you can do this beforehand, it would simplify the analysis.

As it stands now, I recommend the following:

  • LoadImages: Call the image OrigColor

  • ColorToGray: Use Combine to create OrigGray

  • ColorToGray: Use Split to create OrigRed, OrigGreen and OrigBlue

  • ImageMath: Subtract OrigBlue from OrigGreen to get a new image, which we can call GreenOnly

  • SmoothOrEnhance: Use on OrigGray to produce a new image, CorrGray. Use TopHat filtering as your method, and the approximate width of your objects as roughly half that of the object diameter in pixels. If you’re not sure how to do this, look at one of the images in the CellProfiler module windows and go to CellProfiler Image Tools in the window toolbar at top. Select ShowOrHidePixelData and use the cursor to draw a line across a few cells to get a feel for the average diameter (the length of the line is shown in the bottom left corner of the window).

  • IdentifyPrimAuto: Use on CorrGray with Shape as the method to distinguish clumped objects and Distance as the method to draw dividing lines. You can call the resultant objects Cells

Now, the question is: How do you want to calculate transfection efficiency? Do you want to (a) measure the number of DNA objects as compared to the amount of Cell objects? Or (b) do you want to measure the DNA intensity in each Cell object?

Your answer will determine how the image GreenOnly is used. If (a), you can use IdentifyPrimAuto on GreenOnly to find the DNA objects. If (b), you can use MeasureObjectIntensity with Cells as the input object and GreenOnly as the input image, to find the amount of florescence in each Cell.

ColorToGray with Combine will produce a grayscale image with all three color channels with the default name of OrigGray, so you don’t have the option of only using the green channel. In the next module, are you using this OrigGray (or whatever you named it) image? If you can cut and paste the error message, it may help us find out what the problem is.