Hello,
I have data from an experiment in which I'm comparing cell cycle profiles (PI staining) in treated vs untreated cells. The problem was that in certain cases, the number of treated cells was rather low, so I wasn't able to acquire as many cells as in the untreated group. The differences are very clear by looking at the percentages (frequency of parent) of each cell cycle stage (different gates), but I'm quite concerned on how to represent the histograms overlayed. In flowJo, if you overlay the two histograms, it automatically scales the counts' y-axis; if selecting the option "relative to mode" this only changes the y-axis so to be a max of 100; if you select manual however, it plots the actual counts on the y-axis (which are different between my samples). I was wondering if anyone knows what would be the best way to represent this data as overlayed histograms, or whether it is more correct to actually represent them separately, side-by-side, with their actual y-axis counts, and percentage of the population per gate.
Thanks in advance.
Thank you Holger, that was very helpful. I was using "relative to mode", but I'll add in the legend that it shows percentages and not actual counts to make it clear. Best wishes.
Hello, these are cells induced/non-induce with RNAi (starting both with the same number of cells), and the phenotype is that they die overtime, and at certain time points I didn't have enough cells to acquire the same number as the non-induced (let's say 10000 in the induced vs 30000 in the non-induced). I was just wondering how would be the correct way of representing the two histograms overlapped, and I think Holger's suggestion is really good, I'll emphasize that it shows %s rather than counts. Thanks
Holger's suggestion is indeed good. You can add at the m&m the acctual number of accuiered cells, as you will repeat this more than once you can say 20k cells were aquired from the control cells and between 10-13k from theated cells (for example)
Hello Catarina,
In addition to the solution suggested by Holger Fey, I would also like to suggest that, if you want to present data in the form of cell number then you should not throw out the supernatant of your treated cells (collect it in eppendoff in which you collect your treated cells after trypsinization) as they also represent a population of treated cells and you should also prevent the cell loss (even after cell death) by carrying out centrifugation at somewhat high speed (4,000 rpm for 5 min) than normal because dead cells resulted due to drug treatment will be somewhat light. Ideally, you should not be loosing cells as gating only 2-3% of parent population in treated samples will only give you false positive results so if one want to perform the experiment in an accurate number then one must prevent the cell loss. Even there is no use of carrying out analysis with 10,000 cells in control and 2,000 in treatment as instrument will respond differently in two cases.
Also try to resuspend cells patiently in ethanol during fixation as cell clumps can result into shifting of cell population in the debris panel and thus giving a false negative. Make sure that you don't form any cell clumps during fixation as they are hard to break once formed, doesn't matter how much retropipetting you do.
And lastly, you are supposed to acquire the cells at the same voltage and gate conditions in which you have acquired the control cells (that's the whole reason of running a control), if your treated cell's histograms are shifting in control voltage you need to correctly identify the problem (the above suggested solution may handle this).
Hope it helps !
Hello Catarina,
FlowJo v10 has a useful feature for normalizing to the mode called "modal" in the Graph Definition window. This feature will normalize each peak to its mode for each overlayed layer (population). One thing to note is that if "smoothing" is turn on, it might look like each peak is not at 100%, turning off "smoothing" will show all peaks at 100%.
Of course, you can always ask us any question at [email protected], we will be more than glad to assist you!
Best regards,
Miguel V.
FlowJo Tech Support
- Badges
- Science method