The fact that this sounds like it is being taught to a 12 year old. I love it.lol A great teacher. I actually understand the WHYs so clearly now. Thanks.
Hi Aja, great thanks for those clear explanation. My name is Yoo Sun Kim and I'm a postdoc working at NCI(NIH/US). I just started using multicolors flow which is so exciting, and I enjoy learning and using the machine. Would you mind if I ask you a couple of questions regarding compensation controls? I'm looking at tumor infiltrated immune cells populations. 1. Would it be ok if the compensation controls experimental samples are different cells (sources) ? For examples, I use syngeneic mouse splenocyte for compensation controls because cells are positive for most of markers and I get lot of cells. The experimental samples I would like to run is tumor infiltrated immune cells. 2. If the answer from above is yes, would it be acceptable if I use only one marker (such as CD45 which gives strong signal) for all the different parameters? I'm asking because I realized that if I have to compensate markers for active T cells, such asCD69, I had to stimulate splenocyte prior the staining. So I justthought it might be the easiest way if I could use just one marker for all the different parameters. But I was not sure the it makes sense. For example, when I do CD3+ (561nm, YG586) single stain control from splenocyte, I had to decrease the laser voltage pretty lot to see a nice separation, do you think it would it be the same even though I use CD45+ marker for YG586 color compensation? ( I can get the answer if I do it myself :), yes I'm going to order CD45 YG586 antibody..and I'll get back to you when I get the results! but still I just would like to know your thoughts!) Those were two questions I had. Thanks for reading it, look forward to hearing from you. Again, thanks for those fantastic videos! Best, Yoo Sun
Thanks for tuning in and I'm glad the videos have been useful! To answer your questions: 1. yes, this is completely acceptable as long as they cells satisfy the basic rules of compensation controls (just keep an eye on expression levels- some markers can be much higher in the tumor infiltrated cells than in spleen. This would be your primary source of comp errors in this case). 2. for non-tandem dyes, this would be acceptable (though I would use something like CD4 instead that has a clear negative and positive). For tandem markers, you will need to use the exact same vial of antibody due to differences in tandem manufacture and degradation. If you are using this method, you still want to set your voltages based on your experimental samples since this is what you are analyzing/ where you need the best resolution. If your compensation controls are off scale at that voltage, you will need to titrate down the antibody concentration on your comp controls- DO NOT adjust the voltage as this will cause you to lose resolution where you really need it.
Hi Aja, Your video was very helpful. I have run into an assay issue where I use NK and T cells. when I run only the T cells, the populations lies in the first quadrant but when I mix in the NK cells, the T cell population shifts rightwards and spill into the other quadrant. I am wondering if this is a spillover spreading that is causing the shift in population. Would you suggest using curly quadrant gates to fix this?
Hi Aja, I encountered a tough problem on mcherry spillover with other dyes. Here is my case. I started a new experiment to use mcherry labeled lentivirus to edit cell membrane protein X on the cell for knockdown. I detected the mcherry population for viral titration as well as the target protein X expression staining by X-biotin + streptavidin-PE combination due to the low expression of protein X. When I run Fortessa machine to create my compensation for PE(beads), mcherry+ (a mix of transduced cells), unstain (looks like cells are better than beads from this experience). PE was easy to show a narrow peak for positive, however, mcherry was hard to present a peak for positive, which require unstain cells for me to decide a gate to show positive from negative. Even in this careful operation, the machine also warns 200% spillover of PE on mcherry or 200% spillover of mcherry on PE (Sorry, I forget the exact description). How should I do to save this spillover problem? Here is my thinking. First, I will try another dye, such as streptavidin-BUV395, which is not spilled over with mcherry , due to different lasers to excite them individually. Second, if I didn't change fluorochrome, I will try spectral flow cytometry, like aurora. Could you give me a comment or other suggestions? Thanks!
Thanks for the question! I would recommend moving your PE to a different channel (I'm not sure if you are using any other colours... however, there is a lot of overlap between PE and mCherry so you will always get high compensation values with this combination. If you have no other colours in you panel, there is no reason to place them on such spectrally similar dyes, on a conventional cytometer or on a spectral cytometer. On either instrument, you will get a much cleaner experiment by spreading these signals out). BUV395 would be a fine option.
Thank you and glad you found it helpful! While it can be common practice to reuse beads to reuse compensation matrices, best practice is to run them fresh each time. Reuse puts you at high risk for compensation errors and artifacts in your data. So my recommendation is to run them each time with fresh controls.
Thank you for the great explanation. Can you please make a video showing the actual procedure of compensation on your computer? What if you do not use the Wizard but the table? Do you separate the steps, one compensation for the single stains, and another one for the beads? Thank you again.
Thanks for reaching out! I have a separate video on how to setup compensation in FlowJo: th-cam.com/video/bQ6LWA5x0h0/w-d-xo.html When setting up the compensation, you can either use single stain, beads, or a mix of both, depending on what you used for your experiment.
In the top left corner of the FlowJo compensation window there is an optional "Spectral" checkbox. How is that different from what you describe in this video?
The spectral option uses the spectral signature of the fluorochrome to compensate rather than looking at fluorochromes as only existing in a single channel. However, in order for this option to work, you must have enough detectors present and activated to be able to create a spectral signature. For more information, I recommend looking at this: docs.flowjo.com/flowjo/experiment-based-platforms/plat-comp-overview/spectral-compensation/
I would start by trouble-shooting your controls. Make sure that they all follow general compensation rules- especially that they are as bright or brighter than the stain on the samples. This is a very common culprit for compensation issues. And keep in mind that BV750 is a tandem dye so you must use the same antibody to compensate as you do to stain. If all the controls look appropriate (and you have titrated your antibodies), you may want to consider adjusting your panel.
Hi! This means adjusting the compensation manually, using your eye to set it correctly. There are proper ways to do manual compensation; cowboy compensation refers to the “eye-ball it” method of compensating
![Adjusting compensation: how to handle a "bad" matrix [FlowJo advanced]](/img/n.gif)
Great video and explanation! Helped me prep for a job interview doing a lot of flow and this was just what I needed.
Thanks! I'm glad it helped you out 👍
Hello, nice to see your channel! Each episode is very helpful. Thank you for sharing your experience.
Excellent verbal skills. It is engaging with flawless progression throught the entire video. thank you so much!!!
Thanks!!
The fact that this sounds like it is being taught to a 12 year old. I love it.lol A great teacher. I actually understand the WHYs so clearly now. Thanks.
Thanks for the feedback and glad you enjoyed :)
You’re the best, Aja❤ love your videos and I use them as a resource for introducing people to flow.
Yours in Flow, FlowJoe Guy
Thanks! I’ll be releasing some more videos in the near future!
Great video and great explanation
Thanks for the great work.
Hi Aja, great thanks for those clear explanation. My name is Yoo Sun Kim and I'm a postdoc working at NCI(NIH/US). I just started using multicolors flow which is so exciting, and I enjoy learning and using the machine. Would you mind if I ask you a couple of questions regarding compensation controls?
I'm looking at tumor infiltrated immune cells populations.
1. Would it be ok if the compensation controls experimental samples are different cells (sources) ?
For examples, I use syngeneic mouse splenocyte for compensation controls because cells are positive for most of markers and I get lot of cells.
The experimental samples I would like to run is tumor infiltrated immune cells.
2. If the answer from above is yes, would it be acceptable if I use only one marker (such as CD45 which gives strong signal) for all the different parameters?
I'm asking because I realized that if I have to compensate markers for active T cells, such asCD69, I had to stimulate splenocyte prior the staining.
So I justthought it might be the easiest way if I could use just one marker for all the different parameters.
But I was not sure the it makes sense. For example, when I do CD3+ (561nm, YG586) single stain control from splenocyte, I had to decrease the laser voltage pretty lot to see a nice separation, do you think it would it be the same even though I use CD45+ marker for YG586 color compensation? ( I can get the answer if I do it myself :), yes I'm going to order CD45 YG586 antibody..and I'll get back to you when I get the results! but still I just would like to know your thoughts!)
Those were two questions I had. Thanks for reading it, look forward to hearing from you.
Again, thanks for those fantastic videos!
Best,
Yoo Sun
Thanks for tuning in and I'm glad the videos have been useful! To answer your questions:
1. yes, this is completely acceptable as long as they cells satisfy the basic rules of compensation controls (just keep an eye on expression levels- some markers can be much higher in the tumor infiltrated cells than in spleen. This would be your primary source of comp errors in this case).
2. for non-tandem dyes, this would be acceptable (though I would use something like CD4 instead that has a clear negative and positive). For tandem markers, you will need to use the exact same vial of antibody due to differences in tandem manufacture and degradation.
If you are using this method, you still want to set your voltages based on your experimental samples since this is what you are analyzing/ where you need the best resolution. If your compensation controls are off scale at that voltage, you will need to titrate down the antibody concentration on your comp controls- DO NOT adjust the voltage as this will cause you to lose resolution where you really need it.
Hi Aja, Your video was very helpful. I have run into an assay issue where I use NK and T cells. when I run only the T cells, the populations lies in the first quadrant but when I mix in the NK cells, the T cell population shifts rightwards and spill into the other quadrant. I am wondering if this is a spillover spreading that is causing the shift in population. Would you suggest using curly quadrant gates to fix this?
Hi! First question- have you titrated all your antibodies? Normally shifting population backgrounds are due to too much antibody being present.
Hi Aja, I encountered a tough problem on mcherry spillover with other dyes. Here is my case. I started a new experiment to use mcherry labeled lentivirus to edit cell membrane protein X on the cell for knockdown. I detected the mcherry population for viral titration as well as the target protein X expression staining by X-biotin + streptavidin-PE combination due to the low expression of protein X. When I run Fortessa machine to create my compensation for PE(beads), mcherry+ (a mix of transduced cells), unstain (looks like cells are better than beads from this experience). PE was easy to show a narrow peak for positive, however, mcherry was hard to present a peak for positive, which require unstain cells for me to decide a gate to show positive from negative. Even in this careful operation, the machine also warns 200% spillover of PE on mcherry or 200% spillover of mcherry on PE (Sorry, I forget the exact description). How should I do to save this spillover problem? Here is my thinking. First, I will try another dye, such as streptavidin-BUV395, which is not spilled over with mcherry , due to different lasers to excite them individually. Second, if I didn't change fluorochrome, I will try spectral flow cytometry, like aurora. Could you give me a comment or other suggestions? Thanks!
Thanks for the question! I would recommend moving your PE to a different channel (I'm not sure if you are using any other colours... however, there is a lot of overlap between PE and mCherry so you will always get high compensation values with this combination. If you have no other colours in you panel, there is no reason to place them on such spectrally similar dyes, on a conventional cytometer or on a spectral cytometer. On either instrument, you will get a much cleaner experiment by spreading these signals out). BUV395 would be a fine option.
@@ajarieger_flow Thanks for your reply and suggestion!!
Good teacher! Very helpful. May I reuse my compensation if I use the same beads for each single color control? Thank you!
Thank you and glad you found it helpful! While it can be common practice to reuse beads to reuse compensation matrices, best practice is to run them fresh each time. Reuse puts you at high risk for compensation errors and artifacts in your data. So my recommendation is to run them each time with fresh controls.
@@ajarieger_flowI see. thanks a lot!
Thank you so much for these great videos and clear explanations.
Thank you for the great explanation.
Can you please make a video showing the actual procedure of compensation on your computer? What if you do not use the Wizard but the table?
Do you separate the steps, one compensation for the single stains, and another one for the beads?
Thank you again.
Thanks for reaching out! I have a separate video on how to setup compensation in FlowJo: th-cam.com/video/bQ6LWA5x0h0/w-d-xo.html
When setting up the compensation, you can either use single stain, beads, or a mix of both, depending on what you used for your experiment.
In the top left corner of the FlowJo compensation window there is an optional "Spectral" checkbox. How is that different from what you describe in this video?
The spectral option uses the spectral signature of the fluorochrome to compensate rather than looking at fluorochromes as only existing in a single channel. However, in order for this option to work, you must have enough detectors present and activated to be able to create a spectral signature. For more information, I recommend looking at this: docs.flowjo.com/flowjo/experiment-based-platforms/plat-comp-overview/spectral-compensation/
Amazing video, Dr., the link to Flow Post-its from MSKCC is no longer available :((, do you have another link to it?
Here's the new link: wi.mit.edu/flow-cytometry-core/resources/flow-post-its
@@ajarieger_flow Thank you very very much, Dr!!! Learning a lot with your videos, stay safe!!!
Hello- I have a question. What is the best way to correct “over” compensated samples? I have an issue with BV750 and AF647.
I would start by trouble-shooting your controls. Make sure that they all follow general compensation rules- especially that they are as bright or brighter than the stain on the samples. This is a very common culprit for compensation issues. And keep in mind that BV750 is a tandem dye so you must use the same antibody to compensate as you do to stain.
If all the controls look appropriate (and you have titrated your antibodies), you may want to consider adjusting your panel.
hi, I didn't understand what you mean by cowboying compensation, could you explain? :
Hi! This means adjusting the compensation manually, using your eye to set it correctly. There are proper ways to do manual compensation; cowboy compensation refers to the “eye-ball it” method of compensating
Why do compensation matrices change so dramatically over even a single day?
This has to do with drift in the instrument and staining over time
what's with the intro