Mean and Median Fluorescence Intensity - (Sep/29/2009 )
When do you usually espress your data as mean or median fluorescence intensity?
Is it feasible to do this with Foxp3 expression, CD25 and CFSE intensity in a proliferation assay where there's not a good separation of the peaks?
Hi!
Our flow-cytometry lady always suggests to use median, especially if you don't have a nice Gaussian distribution of your cells. Also, median is less sensitive to outliers as compared to mean.
I don't know if it would be good to use with foxp3 or CD25. I would think you would expect that cells are either positive for it or not? Or do you expect varying degrees of expression? For the CFSE it might be possible, but maybe it's better just to look at divided (at all, or highly) as compared to undivided cells? Which programme are you analysing the CFSE peaks with? Flowjo has a function that tries to separate the peaks for you.
Binchen on Oct 1 2009, 04:11 AM said:
Our flow-cytometry lady always suggests to use median, especially if you don't have a nice Gaussian distribution of your cells. Also, median is less sensitive to outliers as compared to mean.
I don't know if it would be good to use with foxp3 or CD25. I would think you would expect that cells are either positive for it or not? Or do you expect varying degrees of expression? For the CFSE it might be possible, but maybe it's better just to look at divided (at all, or highly) as compared to undivided cells? Which programme are you analysing the CFSE peaks with? Flowjo has a function that tries to separate the peaks for you.
As for CFSE, I have CellQUest (Mac) and Summit 4.3 (Windows), but I don't know if they have a function to "deconvolute" peaks. Do you know if they have? Or if there's another good (and possibly free) program that do this? I tried to ask to the SciencesPeak for the CFSE modeler, but it seems that tehy don't provide the program anymore. However, with my cells ( that are cancer cell lines,fairly etherogeneous) I have never had a good resolution of the peaks, so I can't do manually the peak separation.. Maybe you know if has somewhere been analyzed the CFSE proliferation data just with median fluorescence intensity (any reference?)?
As for CD25, usually we find a continuous distribution of the cells,different from what you find for example with CD3,CD4 or CD8 ( like this http://www.leeds.ac.uk/medicine/musculoske...images/cd25.jpg ). Foxp3 is also similar, you don't have a clear separation between what is negative and what is positive, you can do just with a good differentiation with isotype control ( like this http://www.bdbiosciences.com/wcmimages/T_r...mple_data_2.jpg ). So, answering to your question, I would expect various degrees of expression rather than a negative/positive discrimination
thanks for your help