Generally flow cytometric fluorescence data are reported in two ways on a 2D-histogram as indicated in the attached fig.
Some researchers say when there is a shift in almost complete peak (which is rare, generally observed only in control samples, somewhat similar to fig in panel A of attached fig.) then we should use Method 1 and when due to some treatment, a parent population get divided into two subpopulations then we should use Method 2.[In panel B of the given fig main cell population divides into P2( cell subpopulation positive for ROS) and P1( cell subpopulation negative for ROS)]
Can anyone explain, which method is better and WHY? or they can be used interchangeably
Hi Abhinav,
If I have examined the surface markers of my DC's, then with the percentage evaluation. Especially when different points in time were measured. My feeling is that this way, beyond the numbers, I get a better impression of how my population is changing. This applies to surface markers such as CD80, CD86, etc. I cannot say whether this is also true for other endpoints such as intracellular cytokines or similar, but perhaps we can decide differently.
Jochem
If I see two distinct populations by staining, then I would show % of positive populations only to show there is a change in this population. If I see major/minor shift of only one peak by staining, then I would show the quantification of geometric MFI and the overlay of two peaks in the same histogram plot (you can do it in flowjo) to show that there is change in the brightness of total population.
Zhixin Jing Thank you, but what is the reason behind using those two different ways? Why you prefer geometric mean over median/mean for your data.
Abhinav Prasad Please see last response for the reason. We use gMFI over MFI when the distribution is not gaussian distribution. You can also compare two to see how different between them, and which is more representative of the major peak in the histogram.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs