I'm analysing my cells in FlowJo now after having acquired them. I see that '96% of events are on the chart edges'. Does that mean that I have set my FSC and SSC voltages incorrectly?
I acquired about 9 million cells and it's a mixed leukocyte/stromal population so I'm hoping that it's just a lot of stromal cells up there by the edges? Although, when I backgate for CD45 (I'm interested in leukocytes) most of the events are where they should be (but worth mentioning is that they do appear in the top parts of the plot :/ this is at a smaller frequency though)
Thanks for help!
Hi, You could tackle this problem in 2 ways.
1. Go to your original experiment file in diva and activate the biexponential axis. Then export the data as an fcs file and do the analysis in flowjo.
2. If u do not have access to the experiment file, then you could do the biexponential transformation directly in Flowjo. You will find the instructions of how to do the biexponential transformation in the Flowjo website.
When you do this, all the cells that are squeezed in the axis will be clearly visible in the plot.
Hope it helps. All the best!
Hi, You could tackle this problem in 2 ways.
1. Go to your original experiment file in diva and activate the biexponential axis. Then export the data as an fcs file and do the analysis in flowjo.
2. If u do not have access to the experiment file, then you could do the biexponential transformation directly in Flowjo. You will find the instructions of how to do the biexponential transformation in the Flowjo website.
When you do this, all the cells that are squeezed in the axis will be clearly visible in the plot.
Hope it helps. All the best!
If the cells are squeezed on the top edge, then it means that the FSC SSC voltages need to be reduced. The answer that I gave above would work if the cells are squeezed in the X or Y axis.
Best!
Need a bit more information. This is the FSC vs SSC plot, correct? Not FSC vs a fluorescence parameter? The cells can only be “too big” on FSC and SSC, but they can be too small on fluorescence channels due to compensation or baseline restore.
Did you use a BD machine when running the samples? (my experience and comments are applicable to BD/DiVA, not necessarily other machines).
What was the ratio of leukocytes to stromal cells? (if 1:100 then this is obviously predictable/acceptable, but if it is, say, 1:10 then it’s possible that you could be losing blasting leukocytes off the axis).
Also, are you looking at the data on a log or linear scale?
Without knowing these factors I would predict that your FSC is slightly too big and the stomal cells are causing the “problem”. If you’re only interested in the leukocytes then this isn’t really a problem per se, and simply reflects the limitation of the machine to display such a large range of cell sizes on a linear scale. It is possible to simultaneously visualise both small and large cells on the FSC axis if you display the data in log form. However, this usually requires dropping both the FSC voltage and the threshold setting during acquisition.
Hey Isabel,
In order for all of us to best help you, you might include a screenshot of what you are seeing on the graph.
As Ben mentioned, it might be that you are losing cells due to incorrect voltage gain settings for your cell type of interest. FlowJo v10 can change the scaling and range displayed, this might help you, but if the data is out of the dynamic range of detector, i.e. voltages too high, then there might be no fix.
Of course, my colleagues and I are more than glad to help you with your dataset or questions about FlowJo, contact us at [email protected] or here.
Best regards,
Miguel Velazquez
FlowJo LLC Tech Support
I am having a similar issue. During recording on Fortessa I could see all of my populations easily with Diva software. My voltages for ssc and fsc were set so I had a nice even dispersion; however, once I start gating many events are off the page. I'm not sure how to fix it...
Gregory, your problem is addressable. The issue you have is related to baseline restore and the display settings on your Flowjo. The reason the events look fine on DiVA but not FlowJo is because DivA adjusts the display settings to accommodate changes in the negative fluorescence (http://flowjo.typepad.com/the_daily_dongle/2007/02/negative_fluore.html), while in Flowjo this needs to be changed by the user.
You are likely using a default Flowjo width basis of -10, which is fine for most fluorescence channels most of the time. But sometimes, for kind of complicated reasons but usually sample-related, the baseline restore process (a subtraction process that DiVA instruments do on a event-by-event basis) will give a larger negative variance than usual. This will happen on a channel by channel basis. So just because your live-dead looks bad, does not mean your FITC or PE will. This is rectifiable by changing your display settings (some more background here: http://flowjo.typepad.com/the_daily_dongle/2011/09/transformation-differences.html). In FlowJo9, you can change the import settings from a width basis of -10 to -100 (see attachments). This doesn’t affect existing files, just new workspaces. Alternatively, you can change the channels individually (when the data are compensated) in Platform-> Biexponential Transformation -> Manually Specify Transform… (screen grabs for this also attached). Note, however, that changing a width basis from to -100 in compensated data that was imported at -10 can result in a “DiVA spike” due to a binning artefact (http://flowjo.typepad.com/the_daily_dongle/2008/06/diva-spike.html).
This is all much easier to do in FlowJo10, which you have. Just press the big “T” next to each axis on each dotplot, and choose “Customize axis”, which lets you change the display.
You will also find that changing the additional negative display will make your dotplots look a lot nicer.
- Badges
- Science topic
- Similar topics
- Biological Science
- Immunology