Hi Aja! Thank you for your content, it is so nice. So in your opinion, the best flow for optimization of experiment would be to perfom compensation only AFTER voltration (or any other voltage optimization)? Also, what is your opinion about calculating the compensation matrix after acquisition (of both comp tubes and samples), in acquisition software such as diva or even analyse softw such as flow jo? I see a lot of "old" citometry people defending you have to compensate before acquiring your samples (and even doing comp once and saving it for next experiments with same panel). But since it is only a question of math, I feel it doesn't matter - and it is way faster to do it after acquisition. Please, what would you recommend? Thank you once again!
It is a little bit of a chicken and the egg situation and really depends on the individual. Theoretically, if you are starting with titration and then voltration, all your samples are single stained so no compensation is needed. However, some users choose to test the full panel prior to this to ensure the staining and biology make sense- so in this case you may compensate prior to doing the voltration. When and how you choose to compensate is fully up to you and both are acceptable- I routinely compensate in FlowJo after acquisition (even if I have compensated during acquisition). As you say- it is just math. As long as your controls are appropriate and in place it doesn't matter when you compensate so long as you do it before drawing conclusions from the data.
Sorry about the delay in answering- I missed this one! You can add a statistics view and pull the necessary stats after the sample has run and do the calculation on the spot.
Dear Aja, Your videos are really helpful. I did antibody using 1 million cells in 100 ul of staining bufffer. Now, I want to calculate to stain 10 million cells in 500 ul staining buffer for facs sorting. Should I stain my cells for facs sorting with 5-fold higher antibody amount (based on volume) or with 10-fold higher (based on cell numbers)?
Thanks! I'm glad they are helpful. Go based on volume- when staining you want to keep concentration, time, and temperature the same. Cell counts are generally not important in this calculation as, for the most part, you will still be saturating with differences in cell count (unless say you go several orders of magnitude higher). I hope that helps!
How do you recommend approaching setting voltages for FSC and SSC when using cells and beads? For example I have beads for my experimental stains but beads differ in size greatly from my cell type of interest, so I try setting FSC and SSC on cells and then try settin the rest of my comp on beads. Obviously this doesn't always work well because sometimes setting the FSC and SSC via cells will put the beads off scale and vice versa. Any suggestions? (note - I typically use beads for experimental stains because my cells aren't guaranteed to be positive for the experimental stains and therefore they aren't reliable for use in compensation).
This is a common issue. So long as fluorescent voltages remain the same, you can have a set of FSC/SSC voltages for your beads and another set for your cells without issue!
Great video. Question: when I do the voltration, how do I know in which concentration i need to do the process in order to calculate the SI? In other words how is antibody concentration and voltration connected?
Thanks! My general workflow is to start with the antibody titration on a mostly optimized voltage (start with the voltage from the calibration then make sure the positive population is on scale). Once I’ve optimized the antibody concentration, then I go into the voltration :)
Hi Mikhail- I'm glad you found it useful! We use the 8-peak rainbow beads from Spherotech on our instruments. You can see fluorescence in the UV channels; however you likely will not get 8 full peaks in the channels
@@ajarieger_flow Hi Aja I work for a non profit organization promoting flow cytometry. Can I use the leaning tower example for uncompensated data you have shown?? I will promptly acknowledge you
Hi Raja, Thank you so much for the nice presentation. I have a general question for you. I'm going to optimize a flow cytometry method from scratch. Do you think this steps looks reasonable : 1) doing the Voltration. 2) antibody titration, 3) running FMOs and 4) running a sample with all the antibodies. And, would you add viability dye when doing the voltration and antibody titration? If so, how would you deal with compensation? These all looks complicated and dependent on each other and I'm not sure how to start. I'd appreciate your comments on this
Hi Sahar- thanks for reaching out! I would recommend first starting with your antibody titration. This is more critical to good detection (too much antibody and your background will be much larger than it should be; too little antibody and you won't effectively stain your positive cells). Once you have determined your optimal concentration, then go through and do your voltration- you can think of this a the final fine-tuning of your signal to get the best possible resolution. Once you voltages are set, you can then run your compensation controls/ calculate your compensation. Then you're off to the races and ready for all your other samples! (this would be your FMOs and experimental samples. It does not matter what order you run these in.) When you do your antibody titration and voltration, I would recommend just having the stain of interest in your samples, especially when it is your first time going through it. While a viability dye is essential in your experimental samples, you should get a reasonably close stain level in your titration and voltration single stains with out (UNLESS you have a very high amount of dead cells in the single stain samples... at which point you would be better off perfecting your sample prep to minimize dead cells before moving into the titrations and voltrations). I hope that clears things up! Please don't hesitate to reach out if you have more questions!
Hi Aja Thank you so very much again for the thorough reply. It made things more clear in my mind. One more question is: if you start with titration, then how would you make sure that you have an optimal voltage? Would you stick to the CS&T voltages?
@@sahartahvili5081 I would start with the CS&T voltages as a base starting point. After that I would check where the negative and positive populations fall for each stain. So if, for example, the unstained population was really high, I would turn that detector down to maximize the amount of scale on which to detect the positive. After that (and even more importantly), I would run a fully stained sample to ensure that none of my colours was off scale (too bright for the instrument). If any colour is too high, then you would turn the voltage on that channel down. This would be similar to the 10^2 and 10^5 process I talked about in the video. This will give you enough of a base voltage to properly run your antibody titrations.
I use negative beads and positive beads to determine if my antibodies work correctly. The problem I get is that when I go to my brain sample single stained I feel unhappy about recommended voltage as it is not well showing my population I change voltage for one specific channel and then should I replace my bead results with another reading with new setting? I feel beads are way different than my sample. I don't use them for gating either. Can't we just eliminate compensation bead all along? Another weird phenomenon I see is that antibody added to bead show very bright 10^5 signal and on my own sample it is showing a small separate population from unstained below 10^3
Thanks for reaching out! This is a pretty common issue- the antibody capture beads work really well for compensation because they capture so much antibody (and so are almost always as bright or brighter than your samples) but tend to capture a lot more antibody than your cells, depending on the marker. As such, using these beads to set your voltage will rarely result in a voltage that is optimized for your samples. What I would recommend is to set your voltage using your experimental samples and then use those voltages for your compensation. If you are compensating with beads, this may result in some of your bead signals now being off-scale. To correct for this, do NOT lower the voltage. Instead, go and make new beads in for that antibody with a lower concentration of antibody added to the bead (you do not need to use the same concentration on your beads as you use in your samples). If you are compensating with cells, you will have no voltage issues. But you will need to make sure you are acquiring enough events in the file in order for your compensation to be accurate! I hope this helps- feel free to reach out if you have more questions :)
@@ajarieger_flow I see now I will make it into consideration next time it means my antibody concentration for beads need to be less than my sample as it gives super bright result
Awesome video! I was wondering, you mention that your samples should generally fall below 10^5 to stay within linear range on most instruments. How do you typically determine the linear range on an instrument?
Hi Audrey- thanks for the question! That comment is specifically meant for the instruments in my core- they are mostly BD instruments so the scaling ends at 10^5. Other instruments will end at different values... but the same logic applies. You don't want your signal above the highest decade. This paper (pubmed.ncbi.nlm.nih.gov/2582958/) will walk you through the process of determining linear range.
When you adjust the voltage you’re essentially adjusting the sensitivity of detection. When you decrease the voltage, the population goes left (down) because your detection is less sensitive. When you increase your voltage, it goes up (right) because it is more sensitive. I hope that helps!
Dear Aja, Your videos are really helpful. I did antibody titration using 1 million cells in 100 ul of staining bufffer. Now, I want to calculate to stain 10 million cells in 500 ul staining buffer for facs sorting. Should I stain my cells for facs sorting with 5-fold higher antibody amount (based on volume) or with 10-fold higher (based on cell numbers)?
Thank you for the video! Helped me to understand voltage settings!
No problem! Happy it helped 😊
😀 great Chanel #proudtobecytometrist
It was a fantastic explanation! Thank you very much for it
Thank you! I'm glad it helped
Nice video
Hi Aja! Thank you for your content, it is so nice. So in your opinion, the best flow for optimization of experiment would be to perfom compensation only AFTER voltration (or any other voltage optimization)? Also, what is your opinion about calculating the compensation matrix after acquisition (of both comp tubes and samples), in acquisition software such as diva or even analyse softw such as flow jo? I see a lot of "old" citometry people defending you have to compensate before acquiring your samples (and even doing comp once and saving it for next experiments with same panel). But since it is only a question of math, I feel it doesn't matter - and it is way faster to do it after acquisition. Please, what would you recommend? Thank you once again!
It is a little bit of a chicken and the egg situation and really depends on the individual. Theoretically, if you are starting with titration and then voltration, all your samples are single stained so no compensation is needed. However, some users choose to test the full panel prior to this to ensure the staining and biology make sense- so in this case you may compensate prior to doing the voltration. When and how you choose to compensate is fully up to you and both are acceptable- I routinely compensate in FlowJo after acquisition (even if I have compensated during acquisition). As you say- it is just math. As long as your controls are appropriate and in place it doesn't matter when you compensate so long as you do it before drawing conclusions from the data.
@@ajarieger_flow Got it. Thank you so much for those insights! You're awesome!
I just wonder how can I look at the staining Index in the BD machine, right on the spot of doing flow cytometry?
Sorry about the delay in answering- I missed this one! You can add a statistics view and pull the necessary stats after the sample has run and do the calculation on the spot.
@@ajarieger_flow thank you
Dear Aja,
Your videos are really helpful.
I did antibody using 1 million cells in 100 ul of staining bufffer. Now, I want to calculate to stain 10 million cells in 500 ul staining buffer for facs sorting. Should I stain my cells for facs sorting with 5-fold higher antibody amount (based on volume) or with 10-fold higher (based on cell numbers)?
Thanks! I'm glad they are helpful. Go based on volume- when staining you want to keep concentration, time, and temperature the same. Cell counts are generally not important in this calculation as, for the most part, you will still be saturating with differences in cell count (unless say you go several orders of magnitude higher).
I hope that helps!
How do you recommend approaching setting voltages for FSC and SSC when using cells and beads? For example I have beads for my experimental stains but beads differ in size greatly from my cell type of interest, so I try setting FSC and SSC on cells and then try settin the rest of my comp on beads. Obviously this doesn't always work well because sometimes setting the FSC and SSC via cells will put the beads off scale and vice versa. Any suggestions? (note - I typically use beads for experimental stains because my cells aren't guaranteed to be positive for the experimental stains and therefore they aren't reliable for use in compensation).
This is a common issue. So long as fluorescent voltages remain the same, you can have a set of FSC/SSC voltages for your beads and another set for your cells without issue!
Hi Aja- have you covered phosphostainings yet?
Not yet- good idea!
@@ajarieger_flow Would love to watch this!!
Great video. Question: when I do the voltration, how do I know in which concentration i need to do the process in order to calculate the SI? In other words how is antibody concentration and voltration connected?
Thanks! My general workflow is to start with the antibody titration on a mostly optimized voltage (start with the voltage from the calibration then make sure the positive population is on scale). Once I’ve optimized the antibody concentration, then I go into the voltration :)
Thank you! Very clear and simple. Do you know Rainbow Beads working with BD BUV channels?
Hi Mikhail- I'm glad you found it useful! We use the 8-peak rainbow beads from Spherotech on our instruments. You can see fluorescence in the UV channels; however you likely will not get 8 full peaks in the channels
Looks like it is the only available option for today. Thank you Aja for the fast reply.
@@mikhailolferiev7906 good luck with your setup!
@@ajarieger_flow Hi Aja
I work for a non profit organization promoting flow cytometry. Can I use the leaning tower example for uncompensated data you have shown?? I will promptly acknowledge you
@@badrinarayanan3961 Go for it!
Hi Raja, Thank you so much for the nice presentation. I have a general question for you. I'm going to optimize a flow cytometry method from scratch. Do you think this steps looks reasonable : 1) doing the Voltration. 2) antibody titration, 3) running FMOs and 4) running a sample with all the antibodies. And, would you add viability dye when doing the voltration and antibody titration? If so, how would you deal with compensation? These all looks complicated and dependent on each other and I'm not sure how to start. I'd appreciate your comments on this
Hi Sahar- thanks for reaching out! I would recommend first starting with your antibody titration. This is more critical to good detection (too much antibody and your background will be much larger than it should be; too little antibody and you won't effectively stain your positive cells). Once you have determined your optimal concentration, then go through and do your voltration- you can think of this a the final fine-tuning of your signal to get the best possible resolution. Once you voltages are set, you can then run your compensation controls/ calculate your compensation. Then you're off to the races and ready for all your other samples! (this would be your FMOs and experimental samples. It does not matter what order you run these in.) When you do your antibody titration and voltration, I would recommend just having the stain of interest in your samples, especially when it is your first time going through it. While a viability dye is essential in your experimental samples, you should get a reasonably close stain level in your titration and voltration single stains with out (UNLESS you have a very high amount of dead cells in the single stain samples... at which point you would be better off perfecting your sample prep to minimize dead cells before moving into the titrations and voltrations). I hope that clears things up! Please don't hesitate to reach out if you have more questions!
Hi Aja
Thank you so very much again for the thorough reply. It made things more clear in my mind. One more question is: if you start with titration, then how would you make sure that you have an optimal voltage? Would you stick to the CS&T voltages?
@@sahartahvili5081 I would start with the CS&T voltages as a base starting point. After that I would check where the negative and positive populations fall for each stain. So if, for example, the unstained population was really high, I would turn that detector down to maximize the amount of scale on which to detect the positive. After that (and even more importantly), I would run a fully stained sample to ensure that none of my colours was off scale (too bright for the instrument). If any colour is too high, then you would turn the voltage on that channel down. This would be similar to the 10^2 and 10^5 process I talked about in the video. This will give you enough of a base voltage to properly run your antibody titrations.
Hey Aja, Thank you so much again for your reply. You truly helped me alot in understanding and running my control samples.
I use negative beads and positive beads to determine if my antibodies work correctly. The problem I get is that when I go to my brain sample single stained I feel unhappy about recommended voltage as it is not well showing my population I change voltage for one specific channel and then should I replace my bead results with another reading with new setting? I feel beads are way different than my sample. I don't use them for gating either. Can't we just eliminate compensation bead all along? Another weird phenomenon I see is that antibody added to bead show very bright 10^5 signal and on my own sample it is showing a small separate population from unstained below 10^3
Thanks for reaching out! This is a pretty common issue- the antibody capture beads work really well for compensation because they capture so much antibody (and so are almost always as bright or brighter than your samples) but tend to capture a lot more antibody than your cells, depending on the marker. As such, using these beads to set your voltage will rarely result in a voltage that is optimized for your samples. What I would recommend is to set your voltage using your experimental samples and then use those voltages for your compensation. If you are compensating with beads, this may result in some of your bead signals now being off-scale. To correct for this, do NOT lower the voltage. Instead, go and make new beads in for that antibody with a lower concentration of antibody added to the bead (you do not need to use the same concentration on your beads as you use in your samples). If you are compensating with cells, you will have no voltage issues. But you will need to make sure you are acquiring enough events in the file in order for your compensation to be accurate! I hope this helps- feel free to reach out if you have more questions :)
@@ajarieger_flow I see now I will make it into consideration next time it means my antibody concentration for beads need to be less than my sample as it gives super bright result
@@ajarieger_flow Thanks a lot for sharing pearls of wisdom this flowcytometry panicking me everytime
@@auroraprs9133 Exactly!
@@auroraprs9133 if there are any topics that you find particularly challenging, let me know! I'm always looking for new video topics :)
Awesome video! I was wondering, you mention that your samples should generally fall below 10^5 to stay within linear range on most instruments. How do you typically determine the linear range on an instrument?
Hi Audrey- thanks for the question! That comment is specifically meant for the instruments in my core- they are mostly BD instruments so the scaling ends at 10^5. Other instruments will end at different values... but the same logic applies. You don't want your signal above the highest decade. This paper (pubmed.ncbi.nlm.nih.gov/2582958/) will walk you through the process of determining linear range.
Need to know basic about 7 color set up ??
What do you need to know? I would recommend taking a look at our video on panel design :)
when we are adjusting the voltage why the populations moves left right
When you adjust the voltage you’re essentially adjusting the sensitivity of detection. When you decrease the voltage, the population goes left (down) because your detection is less sensitive. When you increase your voltage, it goes up (right) because it is more sensitive. I hope that helps!
I have the quetions❓️
Ask away!
Dear Aja,
Your videos are really helpful.
I did antibody titration using 1 million cells in 100 ul of staining bufffer. Now, I want to calculate to stain 10 million cells in 500 ul staining buffer for facs sorting. Should I stain my cells for facs sorting with 5-fold higher antibody amount (based on volume) or with 10-fold higher (based on cell numbers)?
Sorry I missed this question!! You would go 5 fold higher- always base on the antibody concentration (volume) not on cell numbers