Hi
I have a marker with dim cell surface expression but brighter than FMO but what I am seeing is when I am calculating median fluorescence intensity on flowjo it's giving me a negative value but a positive value for mean fluorescence intensity. Can anyone explain why media fluorescence intensity is negative but mean not?
Hi Ravi,
The median gives you the fluorescence intensity for the median (the "50th percent) cell;, i.e. if you rank your cells by intensity (the least bright first, the brightest last), the median will be the one in the very middle and that value is what you´ll get.
The mean will give you the average, i.e. add all the intensities together and then divide by the number of cells.
In a perfect normal distribution, these values should be identical; however, nothing is perect. Because in flow, we often deal with exponential differences (scaling for fluorescent intensities is mostly in Log-scaling), few very bright cells (or even antibody-aggregates) can have a large impact on the mean, especially if only relatively few cells are recorded in total (I don´t know if that is the case here), which is why median is often considered less prone to outliers. (Imagine 9 random people and Bill gates in a room; the "median" and "mean" wealth in this room will be very different).
As for why you get negative events in the first place; thats due to the compensation. Due to the (somewhat) random emission of photons from a fluorochrome along their predicted spectrum (physics, nothing you can do about it), but a fixed compensation value is extracted from the "raw" value detected by your flow cytometer. Accordingly, some cells appear negative, whilst others appear more positive than they really should. This is why in "light-based" fluoresecnce, you see populations around 0, whilst there is only "0 or more" in mass spec (no comp). Of course, it is also possible that your sample is not properly compensated, resulting in even "more negative" events.
Lastly: Without seeing any plots: Even if the stained sample is brighter than the FMO, if the median is negative, the signal is probably very dim. In fact, it may be so dim that it may also be unspecific binding of the antibody rather than a true signal. If possible, I would try to get a positive control for your marker of interest to confirm the staning.
Hope this helps!
Johannes
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Cell Biology
- Culture Cells
- Cell Line