Which channel is fsw2 on fios

Introduction and summary
On the flow cytometer, different fluorescent dyes are usually used at the same time, the signals of which are measured separately by color filters in the various detectors. Unfortunately, most fluorescent dyes do not emit strictly one color, but a whole spectrum. The fluorescence signal of a dye not only reaches the channel or detector provided for it, but also causes an impulse (albeit a weaker one) in the other channels. This impairs the assessment of the results, since in multi-color analyzes it is no longer possible to know which dye caused the signal from the detector.
However, since one can determine through simple experiments how much a dye scatters into another channel, this can be corrected arithmetically. This process is called compensation in flow cytometry.




Where is the problem?

In the introduction to flow cytometry everything was described quite simply: The lymphocytes are marked with a green fluorescent dye (e.g. CD8-FITC) and a red one (e.g. CD4-PE). The sample is then run through the flow cytometer. The cytometer measures the green fluorescence of the cells in the FL1 channel (FITC channel, 525 nm wavelength), the red fluorescence in the FL2 channel (PE channel, 575 nm).

This gives an ideal separation: the cells marked with FITC are only positive in the FITC channel, those marked with PE are only positive in the PE channel:

Result of staining with CD8-FITC and CD4-PE with correct compensation
The cells colored green in the illustration are CD4 completely negative, those colored red are CD8 completely negative. Just as one would wish, because a CD8 T cell does not express CD4 and a CD4 positive T cell does not express CD8.

Note: the green or red coloring of the cells is arbitrary. It only serves to highlight the populations.

But that only works because the measured values ​​of the cells are mathematically corrected, they say "compensated". Without this computational compensation, the plot would look very different:

Result of staining with CD8-FITC and CD4-PE WITHOUT compensation
This is what it looks like without compensation. This is what the real readings look like that the flow cytometer measures. There are practically only doubly positive cells: all cells are more fluorescent in both channels than the negative cells (black).

This plot looks less beautiful. Our CD8-positive cells (in green) all also seem CD4-positive. And our CD4 positives (in red) seem to be CD8 positive too.
However, this does not correspond to the biology of the cells, because the CD8-positive should not be CD4-positive and the CD4-positive should not be CD8-positive (at least in the blood). The uncompensated display leads to strange results.

To simplify matters, we only stain with CD8-FITC, then there shouldn't be any double-positive cells, because we didn't add any "red" antibodies:

Result of staining only with CD8-FITC WITHOUT compensation
Although we do not add any "red" antibodies at all, the CD8-positive cells also appear positive in channel FL2 (the "red" channel).

This plot doesn't look any better either. All CD8 positive cells are also positive in the second channel. Why is that? In front of the detector of the second channel (FL2) there is a color filter that does not let green through at all.


Cause: FITC doesn't just glow green
In our idealized imagination, a FITC-positive cell only glows green. Because FITC is a green fluorescent dye. Unfortunately this is not the case. FITC doesn't just glow green. Although it shines green the strongest, it also shines at many other wavelengths.
The wavelengths emitted by FITC can best be represented with a wavelength or frequency spectrum. Since it's about the charisma, that's what it's called Emission spectrum.

It looks like this at FITC:

FITC emission spectrum
FITC not only emits green (peak), it also emits longer wavelengths. FITC also emits yellow and red, albeit much weaker.

The spectrum shows: FITC also shines in other colors. A FITC-positive cell will therefore also trigger an impulse, albeit a smaller one, in other channels of the flow cytometer.
You might say now that this minimal amount of red could not have made such a strong signal in FL2 after all. The FITC-positive cells were almost in the middle of the FL2 axis, around 300.
To understand this one has to free oneself from a prejudice: PE does not fluoresce red but yellow and the FL2 / PE channel therefore does not measure red fluorescent light but yellow. And the yellow component of FITC is considerable, as the spectrum above shows.
Background info: The misconception of the "red" PE can probably be explained by the development of flow cytometry. Initially, flow cytometers only had one fluorescence channel, mostly for FITC, the green dye. Then a second dye was added - PE. Since red-green representations are much prettier than red-yellow, the PE was often symbolized with the color red. The antibody manufacturers also filled their FITC antibodies in green bottles and PE antibodies in red bottles. This created the wrong picture of the red PE.

Now it becomes clear why the FITC-stained cells look like this in the flow cytometer:

The considerable yellow component of the FITC fluorescence is also measured in the PE channel. That's why the CD8-positive cells are so high up.

Is that even a problem?
Yes sure. For example, CD4 / CD8 double-positive lymphocytes are very likely to occur in certain lymphomas. In fact, this is of great help in diagnosing these diseases. To recognize this, one could stain the lymphocytes with CD8-FITC and CD4-PE at the same time. But if the CD8 cells are already strongly positive in the PE channel because the CD8-FITC radiates into the PE channel, then it is very difficult to see whether the CD4-PE is also on the CD8-positive cells .




How do you solve the problem?

The problem is solved with a mathematical correction: If I know that FITC also radiates into the PE channel, then I know that I have to subtract something from the PE signal that I am measuring. Namely, exactly the portion that comes from the FITC.
There is usually a special menu in the flow cytometer software for this purpose. There you will find information that might read:

FL2 = FL2 - X% FL1

analogously it could also read:

FL2true = FL2measured - X% FL1measured

This means that the FL2 signal is calculated as follows: a certain proportion of the signal in FL1 (the FITC signal) is subtracted from the measured value in channel FL2. In other words, the portion of the FL2 signal coming from the FITC is subtracted from the FL2 signal. This gives the true level of the cells' FL2 / PE expression.
This process is called the Compensation.

How high the percentage that has to be deducted is either determined by the device in an automatic compensation routine, or you do it yourself, which is very informative.


How do you set the compensation?

The following is an example of how to set the compensation for FL1 against FL2.
The lymphocytes were stained with CD8-FITC:

This is how you go about it practically

Similar to what is shown in the figures, you can actually set the compensation. To adjust the compensation, a sample with two cell populations is required: one population of unstained cells and one population of colored cells. Otherwise, the cells of the two populations should be as similar as possible. Usually you choose lymphocytes (by selecting them in the scattergate). Appropriate markers can be used so that some of the lymphocytes are stained while others remain unstained. E.g. CD4, which only stains part of the lymphocytes. It is even more correct if you first prepare 2 samples. With one sample, the lymphocytes are stained, e.g. with CD45, with the second sample, one does not stain. Before the measurement, the 2 samples are poured together. You now have a CD45-positive, colored and a CD45-negative, unstained population in the sample. The two populations are otherwise completely identical.
Monocytes and granulocytes are not so well suited for setting the compensation due to the higher level of autofluorescence.


Do you have to set the compensation so precisely?
That would be ideal. The problem is not so great when assessing highly positive and clearly deposed populations. But an assessment of weak antigen expressions is difficult if the setting is wrong. Even with continuous antigen expression ("smeary" expression) wrong results could arise.




PE also radiates into other channels

It's not that only the FITC fluorescence scatters into other channels. The PE also not only emits yellow, it emits at different wavelengths, as its emission spectrum makes clear:

Emission spectrum of FITC and PE
PE not only emits yellow (peak), it also emits shorter wavelengths. It also shines a little green. A PE signal is therefore also detected in the FL1 / FITC channel.


Therefore, even a sample colored with CD4-PE does not look optimal without compensation:

CD4-PE without compensation
The green component of the PE fluorescence is also measured in the FITC channel. That's why the CD4-positive cells are so far to the right.


As you can see, you don't have to compensate too much from PE to FITC. This is mainly due to the fact that PE only has a very small proportion of green (see PE emission spectrum above). Still, you have to compensate a little. Of the above options, a compensation of 0.5% appears to bring the best result. 0.3% is still undercompensated and 0.7% is already overcompensated. At 0.5%, the population of CD4-positive lymphocytes is almost exactly above that of CD4-negative cells (shown in black).


Can overcompensation be a problem?
Yes, often a bigger one than the undercompensation. If you undercompensate, you may mistakenly think of a population as positive. If you overcompensate strongly, you may consider a weaker positive population to be negative or, in the case of strong overcompensation, you may miss the entire population because it is pressed against the axis (see example above, button 5).


Technical Notes

  • Sometimes you would want an additional decimal place when setting the compensation, because the ideal value is e.g. between 0.4% and 0.5%, i.e. maybe 0.45%. This is not possible with all flow cytometers. Usually this is not so important, because the compensation fluctuates a little from antibody to antibody anyway and therefore a too precise setting with a certain antibody is of little use.
  • In addition to the optical assessment, a region is of course placed over the positive and negative population and the median fluorescence intensity or the geometric mean of the fluorescence intensity can be displayed. These values ​​should be about the same for both populations.


For advanced

Knowledge of the following points is not necessary to set a correct compensation, but the questions may be of interest to you.


Why high compensation is not a good thing

After reading the above, you might think that it doesn't matter if two fluorescent dyes that you are using overlap a lot. You can compensate for the whole thing arithmetically. In principle yes, but it is still not recommended to work with very high compensation factors. Not only because a slight incorrect compensation would then have major consequences, there are also disadvantages with optimally set compensation:

Measurement errors are carried over to the other channel
Let us first imagine the ideal case: a sample is stained with e.g. CD45-FITC. You have determined that the compensation between FITC and PE must be 40%. A cell has a measured FITC signal of 100 and a measured PE signal of 40. You know that you now have to subtract 40% of the FITC signal from the PE signal. 40% of 100 is 40. The calculated PE signal is the measured minus the FITC component, i.e. 40 minus 40. Results in 0. Ideal.
In reality, a measurement is never error-free, and it is possible that the FITC signal was incorrectly measured at 90. The PE channel measured correctly, i.e. 40. 40% of 90 are only 36. If you now subtract 40% of the FITC signal, i.e. 36, from the PE signal, there is still 4 left over. So the cell seems to have a PE signal. But only because the measurement in the FITC channel was incorrect.
On the other hand, if too much is incorrectly measured in the FITC channel, the PE signal is incorrectly compensated too much (in our case negative values ​​would result, which is not possible. But if the cell had a real PE expression, it would then corrected too much downwards by the compensation).

The precision in the adjacent channel decreases
In total, the effect leads to a greater spread of the PE values.
If the fault in the FITC channel is a random fault, which is to be assumed. A systematic one should have been taken into account when setting the compensation or it should have been noticed during the analysis.

Lymphocyte population on the left without antibodies, on the right stained with CD45-FITC:

Do you see the difference in the "PE expression" of the two populations? Maybe in the enlargement:

Lymphocyte population on the left without antibodies, on the right stained with CD45-FITC:
On the right-hand side, the population in the PE expression spreads somewhat more. Both down and up. This is caused by the transmission of the measurement error from the FITC channel to the PE channel. As described above, this is transferred through the compensation.

Admittedly, the effect is not particularly great because the measurement errors of modern devices are usually not very high. But the higher the compensation and the greater the measurement error, the greater this effect.


A common mistake

As explained in the previous paragraph, the positive cell population may spread a little more, i.e. have a wider distribution range than the negative. Yes, it really has to be like that. This means, however, that even if the compensation is set correctly, a few cells of the positive population fall above the threshold that was set after the negative control. That's OK. It would be wrong to increase the compensation until no more cells are above this threshold. An example may make this clearer:

Analysis of a mixture of CD45-FITC-stained lymphocytes and unstained lymphocytes.
The compensation is set quite well, the two clusters are about the same height. Because the CD45-positive cells spread more, a few are above the threshold (arrow). But that shouldn't bother you. That's right.
Analysis of a mixture of CD45-FITC-stained lymphocytes and unstained lymphocytes.
The compensation has now been set higher so that no more cells are above the threshold. However, this sample is clearly overcompensated: the pile of CD45 positive cells is lower than that of negative cells.



Why do you have to change the compensation when the PMT voltage changes?

You have correctly set the FITC-PE compensation. Then you change the PMT voltage of your FITC-PMT and suddenly the compensation is no longer correct.
At first glance, one might wonder why? The yellow component of FITC is always the same, no matter how high you set the PMT voltage.
Yes, the proportion is always the same. But the reading you get will change after changing the PMT voltage. And if you just change the FITC or just the PE-PMT voltage, then the value you have to compensate with also changes.

They had optimal compensation at 20%. This means that if you measured an FITC signal of 200 in a sample that was only FITC-stained, you measured 40 in the PE channel. You subtracted 20% of 200 so 40 from the PE signal, that resulted in 0 and everything was fine.
If you now reduce the FITC-PMT voltage, you may only measure 160 in the FITC channel. Not because the cell fluoresces more weakly, but only because you have set the FITC channel PMT amplifier to a weaker level. However, you have not changed the PMT amplifier of the PE channel, so you will continue to measure a signal of 40 in the PE channel. A 20% compensation is no longer sufficient. 20% of 160 is 32, if you subtract 32 from the PE channel signal of 40, there are 8 left. The sample is undercompensated.
If you decrease PMT voltages, you have to increase the compensation against the adjacent channels; if you increase the voltage, you have to decrease it.