Hi @biscotrash - serial dilution is the way to go, it makes it much easier to handle the lower concentrations of your standard curve!

Hi Hana, found a new position in industry and no longer works at the UoC. We'll be happy to help - find us on google and follow the contact us link.

@@UChicagoFlow Thank you so much for replying. I tried but I was not able to find you.

@@hanaabdullah6662 voices.uchicago.edu/ucflow/contact-us/

@@UChicagoFlow please check linkedin

Hi! I'm guessing this happening on a Canto? I haven't seen one of these in years, by my guess would be an issue with the sheath pressure not getting high enough to go through startup. I'd check the pressure gauge on the fluidic cart, purge any air stuck in the fluidic line, your tubing connectors for any leaks. Again, wild guess here! Can't be sure without having a look at the instrument.

You could overwrite the references. Or it may be better to add new tubes in the reference group so that you have all the options when unmixing the data. It's easiest to do this on the actual aurora - if you ever choose to do unmixing on another computer you'll have to move some instrument settings as well. Depending on how complex your experiment is, you might find it helpful to discuss with your Cytek Technical Applications Specialist to advise you on the best approach for your specific experiment.

@@UChicagoFlow Thank you for the stunning presentations! Please, how to transfer instruments settings between computers? I'm planning to unmix/analyze data on another computer to save machine time. Thanks!

@@pavlabohacova5906 You'll need administrative rights to make this change, so it may be best to talk to whoever maintains your Aurora. You'll need to transfer instrument settings using the data maintenance tool. Also my understanding is that you only need to do this if you're unmixing files that were recorded on different days. If everything was run on the same day you just need the fcs files/experiment.

Hi Fei, thanks for the good word. Unstained beads will be used as control for compensation and unmixing if you are using the same beads for your positive group. The autofluorescence of both positive and negative must be the same. But for your FMOs, the question is different: you are trying to figure out the impact of all of your fluorophores (minus the one you're gating on) on the spread of the negative population. So the only appropriate way to create this control is to use cells.

@@UChicagoFlow thanks for your response. I am still confused with the unstained control. In my previous experience, I used unstained beads to adjust FSC and SSC to settle the cell population in the middle of plot. And set up the voltage of each color through the single control with beads for compensation. However, I didn’t do unstained cells to adjust that cell population. Beads could not the same size with the real cells. So, I want to know whether my design is acceptable through unstained beads to gate cell population instead of unstained cells? Sorry for my tedious words.

@@feitu6403 I typically set voltages on FSC and SSC specifically for the beads and cells. Most of the time, they have different properties, and will require you to decide on the settings to best showcase what you're actually looking at. As far as selecting the voltages for the fluorescent parameters detectors, I think your method likely puts you within range of what will allow you to separate your positive and negative markers without saturating the detectors. This is the whole point and if that is what you have right now, you are probably good to go. We go in more details about voltage selection in the next video in this series - here 13:05.

@@zxcv2677 thanks a lot!

When creating or editing your experiment, on the group tab select create or edit reference group, then in the window that lists all of the reference control tubes there is an "Add" button in the lower left corner that will allow you to add as many reference controls as you need.

Hi Claudia - Good catch, we'll figure out a solution on accessing the link on the slide. It's on the CAT Facility Blog: voices.uchicago.edu/ucflow/2020/01/16/how-many-cells-should-i-stain-the-impact-of-cell-concentration-and-antibody-concentration/

I would not recommend combining the beads and cells into a single tube two two reasons. First you would need to be 100% certain that the beads and cells do not overlap on the FSC vs SSC plot. Depending on the tissue used, there could be some overlap. Inability to separate beads and cells could be problematic for how autofluorescence is accounted for in terms of calculating unmixing from the reference controls. Second is that beads and cells don't necessarily require the same concentration of antibody for optimal results. I could come up with a bunch of advanced tricks that would allow you to combine beads and cells into the same tube, but ultimately it's going to take a lot more effort than it's worth.

Personally, I have not found that including a viability stain for titrations was necessary. If you do have a high percentage of dead cells then you may find it beneficial to include a viability stain. I don't think people typically add in the viability stain with titrations unless they're struggling to interpret the data without it. The extra tools menu will let you unmix any number of fluorophores, just like the unmixing wizard in the acquisition tab.

@@UChicagoFlow Thank you very much for your expertise! Really highly appreciated! If I dare to ask, I would have 1 question concerning wet lab work. For the final staining, we would use 1 x 10^6 cells in 100 µl. For the titration, we would like to use less cells, eg. 100.000. Would you suggest to keep the volume constant (use 100.000 in 100 µl) or would you suggest to use less volume. Of course we would keep the final antibody concentration constant. I would really appreciate your help, since the literature on this topic is not quite clear!

@@Teresa-xt8oi There is a blogpost on the CAT Facility website with data supporting the idea of keeping the volume the same for your controls : voices.uchicago.edu/ucflow/2020/01/16/how-many-cells-should-i-stain-the-impact-of-cell-concentration-and-antibody-concentration/

@@chugcytometry3284 Thank you for your help!!

I'm sorry, I'm not familiar with this cell type so I don't think I can advise you. I don't think cell strainers are an absolute requirement, but they work well for cells that are often used in flow cytometry. You'll have to find a method of single cell preparation that is appropriate for your cell type.

@@UChicagoFlow Thank you for taking your time to reply. I appreciate it. I will continue to search for the optimal conditions.

Yes, it is possible. Just to make sure we're all on the same page: unmixing can be done in multiple applications (SpectroFlo, FlowJo, FCS express), though I personally find SpectroFlo the most user friendly. With SpectroFlo unmixing there is no way to view the "unmixing matrix". If you want to make a manual adjustment, you must use a separate method - compensation. Compensation can be done in multiple applications (SpectroFlo, FlowJo, FCS express) so that compensation is essentially applied on top of unmixed files. I would recommend manual compensation, not an automated compensation wizard. I am less familiar with FCS Express, but in FlowJo you must open the compensation editor and click "edit", which will generate and apply a new compensation matrix. To get a negative comp value you can select a box in the matrix and hit the down arrow key - typing a negative sign doesn't work but once you have a negative value you can manually type any value after the negative sign. And I feel the need to reiterate the most important point: manual compensation should be used with caution and for data that will be presented/published you are only allowed to adjust the compensation to correct the single stain controls and then apply that to the fully stained sample. Never ever adjust compensation directly on a fully stained sample.

Thank you very much! Which topic would you like us to cover next?

@@UChicagoFlow Please upload a detailed video of system calibration , standardisation and laser delay.

The overlooked point about the Spectral flow instruments is that you can, as you mention, just use the raw data without unmixing. You'd essentially be using the instrument as a traditional flow cytometer, and that is totally fine. The width of the bandpass of each detector is usually much smaller than what you'll find on you typical flow instrument, so fewer photons will hit the detector. But if the signal is bright enough, it should not matter.

@@chugcytometry3284 Thank you for the helpful response! I just titrated antibodies using mouse lung and was worried why a few populations hadn't resolved (CD206 BV605 and Siglec-F BV510). I'm hoping that it's a brightness issue that will be resolved once I go back and properly unmix the files this week.

This question comes up a lot, and I don't think theres a consensus on which way is best. I think a lot of people end up trying both ways.

@@kaylaparr7732 Do you mean that you don't have a positive signal at all, or that you can separate your positive cells from the autofluorescence?

@@chugcytometry3284 Separating positive cells from autofluorescence. There is positive signal in all serially-diluted antibody samples but no clear positive and negative populations (all tubes, including unstained, look like a lot of fluorescence spread).

If you just use spectral unmixing in SpectroFlo and do not use any compensation in addition to that, then the "uncompensated" and "compensated" parameters in FlowJo are the same. If you did apply compensation on top of the spectral unmixing (either in SpectroFlo or FlowJo) then you'll want to choose the compensated parameters in FlowJo. It's best to do a panel quality check in data that is free of any unmixing errors.

You're very welcome!

Hi Lazar, Generally, I would say no. Most people titrate antibodies as a single stain. Doublet discrimination is important for ensuring that we're not getting false positives. Imagine if two single positive cells are stuck together - without doublet discrimination we might think we have a double positive cell. Because there's only one antibody to titrate in one tube the doublets aren't really a concern. Viability is also important for accuracy of your experimental data, your antibody might be nonspecifically binding to dead cells. See this blog for further explanation: voices.uchicago.edu/ucflow/2020/11/19/how-to-identify-problems-with-panel-design-bad-data-part-2/. For titrating an antibody the dead cells usually don't interfere too much with determining the optimal concentration. But I guess if you're seeing a much higher frequency of positive cells than expected or it seems like the background is really high maybe it's worth it to include a viability dye. I've yet to run into that situation, but I wouldn't say it's impossible. In that case you'd have to then worry about compensation between the antibody and viability dye, which is just more effort than a single stained tube. The simplest experiments are ideal.

Hi! This can be a tricky question. First I want to point out that the frequency and intensity of a marker are distinct features and independent from each other. You can have a low frequency population (<1% of the total cells) that expresses 100,000 molecules per cell (high expression/high intensity). Alternatively, you can have a high frequency population (maybe 60% of the total cells) that expresses 2,000 molecules per cell (low expression/low intensity). For a single stained, control the fluorophore intensity of the control must be as bright or brighter than the fully stained sample - regardless of the frequency of that positive population. If the positive population is rare, then you may need to record more total events in order to get enough positive cells, but this does not affect the intensity of the marker. If it is difficult to get enough positive events (100 would be the absolute minimum, but more is better), you may consider using compensation beads. My best recommendation is to stain the single stained controls and fully stained samples in the exact same way - tissue type, cell concentration, antibody concentration, incubation time, incubation temperature. The easiest way to mess up the controls is if people use one tissue type for their controls and another for their fully stained tube and it turns out that the tissues have different levels of expression for a marker. Or mistakes can be made in pipetting the single stained controls - especially when pipetting volumes less than 1 uL.

Each time a tandem dye is created, there may be slight variations in the signature. Meaning the signature from lot 1 may be slightly different from lot 10. Because the signature of the dye in the control must be the same as the signature in the sample, you must use the exact same vial of antibody to stain the control and the sample. If you use a different vial of antibody (such as a different lot or an entirely different antibody with the same fluorophore), you will likely not have the same signature in your control and sample and will end up with unmixing errors. Non-tandem dyes do not have this lot to lot variability and so you can switch antibodies or lots for those fluorophores.

@@UChicagoFlow Yes, Indeed we can prepare the BV750 reference control from the same antibody lot that is going to be used for the sample. There is always a slight variation in the staining pattern/signature in tandem dyes from lot to lot due to fluorescence/protein ration, the pairing of tandem molecules/fluorophores etc.

Hi! Rule 5 states that the signature of the control must be the same as the signature of the sample. Fixative can slightly alter fluorophore signatures, so if your controls are unfixed and your samples are fixed the signatures will not be the same and you will get unmixing errors. You must either have both controls and samples fixed or both controls and samples unfixed.

@@UChicagoFlow It means to say that the unmixing is different from conventional compensation where it says that the compensation control should be brighter than your sample control. So, for the unmixing algorithm/feature, we always need to use a similar treatment or conditions for references and samples to get proper unmixing results. As in the compensation rule, it says that the background fluorescence should be the same for negative control and positive control (Means if there is a treatment in your positive control/reference then the similar treatment should be given to negative control/reference, no comparison with the sample condition/treatment)

@@prateekarora414 Actually, the rules to create proper single stain controls are the same - whether you decide to use them to perform spectral unmixing or compensation. Since compensation can be performed manually, people are often able to compensate their data even if their single stain controls do not follow all of the rules. And because conventional cytometers only look at a fraction of the overall emission spectra, they may not be able to detect lot-to-lot variability as well as spectral cytometers, so using a different lot of antibody between controls and samples may not have as big of an impact on compensating data as it does on spectral unmixing. Your comment also referred to a "compensation rule" - I'm going to assume you meant the third rule in my video. I'm not sure if you understood this rule correctly and what you mean by "treatment" (fixation? stimulation? disease model vs control?). It's challenging for me to provide detailed explanations in KZfaq comments - if you have questions about what controls to use for your experiment, I would suggest contacting your Cytek Technical Application Scientist and they should be able to discuss this with you.

​@@UChicagoFlow Thanks and appreciate your valuable time to give such commendable comments. I do apologize if, my comments annoyed you. I completely understood the concern of rule 03, where it says that the autofluorescence of the negative population in the positive control should be identical to the autofluorescence of negative control (Highly recommended in spectral cytometers to matching the background for better unmixing) else use universal negative in third party software. I mean of treatment was for buffers only means as you said in rule 05 that all the controls should be treated with the same buffers as your sample to get the identical spectral signature but if someone is using beads then I presume there is no need of treating the beads with the same buffers as for the samples, I know the treatment will alter the fluorescence spectrum but here if we add the treatment buffer like a fixative buffer to the beads references controls then the spectral signature would not be similar to the cells treated with same fixative buffer..

@@prateekarora414 I'm sorry if I sounded annoyed, that was not my intention! I just want to make sure you're getting the answers to all of your questions and I am not able to answer everything in full detail using KZfaq comments. Please contact your Cytek Technical Application Scientist and they'll be able to have a discussion with you and answer all of your questions.

Thanks, that's good to hear! Let me know if you have questions!