Hi Dr, Rieger. First, thanks for your channel, it has clarified a lot of questions I had. I am wondering if the gate adjustment you showed (min ~8) can be done when using unstained control (?). Digging on my data, I checked the comp area and I had to adjust the axis to enlarge the peak and only then I was able to see 2 peaks (I assumed, the positive and negative peaks). Is that expected when using unstained control for compensation? In your video boths peaks are pretty clear from the very beggining, but I used unstained control and I had to adjust the axis and then adjust the gating. Is that correct? I would appreciate your help!
I wouldn't recommend adjusting compensation with the unstained control as almost all compensation issues (over and under comp) will be seen in the positive population. The split negative is usually due to the scaling. I would recommend taking a look at this: www.omiq.ai/guides/understanding-data-scaling
Thank you very much for your videos. I have rather a simple question and would greatly appreciate any advice. What do you use for your dead cell exclusion? I am confused because I use 7-AAD and it spills over to my other channels of TMRM and MitoSOX Red dyes, hence I can't exclude dead cells prior the compensation matrix formation. I can't use other viability dyes because they overlap even worse with my other dyes in the panel of 5 dyes.
I would have a compensation control for 7-AAD as well (or any other viability dye). You don’t have to exclude dead cells prior to compensation matrix formation but you do need to account for the fluorescence from the dye you use.
Well said, lot of good points. I will suggest users to bring the negative population from a completely unstaned bead or cell. The negative bead or cell in the single stain will have some level of NSB that will lead to violation of rule of AF. DIVA only allow one universal negative but using more than one universal negative will be a good idea (cells for LD stain and Beads for Ab stain) for analysis.
What if the cell population cannot be separated well within a single color control for compensation? How to identify negative and positive peak in this single color for compensation? Thanks!
Sorry- I have gotten behind on my responses! In that case, because compensation doesn't care about biology (i.e. if your gates are in the right spot to capture the correct population) but only cares about the fluorochrome, you can just take the extremes of each side (the most negative and the most positive) and use that. Just make sure you have enough cells included!
I made the mistake of reusing my compensation matrix because I have a very large panel (18 colors). My matrix only had one major spillover issue of 153% but the outcome is that it is showing my TCRb+ T cells expressing IgG's :') everything else looks fine, its even been looked at by Flowjo people (briefly) :( I can't redo it either as I did it on brain cells and have nothing banked (near impossible to bank brain cells after stroke). I am trying to graduate in the spring and I am STRESSED.
What if the issue is literally just those 2 fluorophores with each other? Would you just ignore those as not trustworthy but everything else be ok?@@ajarieger_flow Thank you for the response, I am so lost haha! I am doing clustering analysis and UMAP so I have clusters of TCRb+ IgG+ T cells. The whole graph just rockets.
![FlowJo [UMAP]](http://i.ytimg.com/vi/fuTOiaPr86o/mqdefault.jpg)
![FlowJo [UMAP]](/img/tr.png)
![FlowJo [tSNE]](/img/1.gif)
Love your videos! So helpful! Thank you.
So helpful! Editing the matrix has always been my go-to to fix a bad compensation. 🤐
You’re not alone- many a matrix has been manually edited. Glad this helped!
Hi Dr, Rieger. First, thanks for your channel, it has clarified a lot of questions I had.
I am wondering if the gate adjustment you showed (min ~8) can be done when using unstained control (?). Digging on my data, I checked the comp area and I had to adjust the axis to enlarge the peak and only then I was able to see 2 peaks (I assumed, the positive and negative peaks). Is that expected when using unstained control for compensation?
In your video boths peaks are pretty clear from the very beggining, but I used unstained control and I had to adjust the axis and then adjust the gating. Is that correct?
I would appreciate your help!
I wouldn't recommend adjusting compensation with the unstained control as almost all compensation issues (over and under comp) will be seen in the positive population. The split negative is usually due to the scaling. I would recommend taking a look at this: www.omiq.ai/guides/understanding-data-scaling
Thank you very much for your videos. I have rather a simple question and would greatly appreciate any advice. What do you use for your dead cell exclusion? I am confused because I use 7-AAD and it spills over to my other channels of TMRM and MitoSOX Red dyes, hence I can't exclude dead cells prior the compensation matrix formation. I can't use other viability dyes because they overlap even worse with my other dyes in the panel of 5 dyes.
I would have a compensation control for 7-AAD as well (or any other viability dye). You don’t have to exclude dead cells prior to compensation matrix formation but you do need to account for the fluorescence from the dye you use.
@@ajarieger_flow Thank you very much for such a fast reply!!!
Well said, lot of good points. I will suggest users to bring the negative population from a completely unstaned bead or cell. The negative bead or cell in the single stain will have some level of NSB that will lead to violation of rule of AF. DIVA only allow one universal negative but using more than one universal negative will be a good idea (cells for LD stain and Beads for Ab stain) for analysis.
Thanks- and yes, decent point for those using a universal negative.
What if the cell population cannot be separated well within a single color control for compensation? How to identify negative and positive peak in this single color for compensation? Thanks!
Sorry- I have gotten behind on my responses! In that case, because compensation doesn't care about biology (i.e. if your gates are in the right spot to capture the correct population) but only cares about the fluorochrome, you can just take the extremes of each side (the most negative and the most positive) and use that. Just make sure you have enough cells included!
Thank you soo much .. :)
I made the mistake of reusing my compensation matrix because I have a very large panel (18 colors). My matrix only had one major spillover issue of 153% but the outcome is that it is showing my TCRb+ T cells expressing IgG's :') everything else looks fine, its even been looked at by Flowjo people (briefly) :( I can't redo it either as I did it on brain cells and have nothing banked (near impossible to bank brain cells after stroke). I am trying to graduate in the spring and I am STRESSED.
Oh no!! Good luck! And remember that over 100% is not necessarily wrong. Look at you nxn plots to see where you issues lie
What if the issue is literally just those 2 fluorophores with each other? Would you just ignore those as not trustworthy but everything else be ok?@@ajarieger_flow Thank you for the response, I am so lost haha! I am doing clustering analysis and UMAP so I have clusters of TCRb+ IgG+ T cells. The whole graph just rockets.