Hi Samarjit

Suppose that you standardize the flow cytometry protocols. Regarding

Flourochromes , gain , voltage , acquisition , compensation

Theb you will not face any differences

Dear Samarjit Maharana,

For example, imagine that you would like to compare CD4 expression between samples obtained with different flow cytometers.

Then, you are basically allowed to compare MFI (mean fluorescence intensity) data between samples analyzed in the same setting & the same machine.

However, if each of your exps has included an appropriate control to correctly measure the CD4+ frequencies (%), then you would be able to compare CD4+ frequencies among all the samples tested regardless of the used machines.

You could also be careful about the gating strategies (FSC/SSC, etc) because these could affect MFIs as well as frequencies of the populations of your interest.

Thanks Dear Tsukamoto,

In Canto-II I have some of my control run with patients and I want to check the % different between them and same type of experiments I have in Canto-I for another set of control and patient.

So, as per your suggestion there will not be any problem to use all my data to compare. Is not it?

Thanks

The answer depends on whether your controls are appropriate for your purpose. If these are isotype controls using the same antibody, then I guess yes.

If you mean "healthy volunteers" as controls and you have included different samples for different analyses, the intended comparison could be uncontrolled.

Anyway, it depends on the design of the study, i.e. machine-independent controls should be designed prior to your analyses.

Yes, I am using healthy controls for each experiment along with patients sample.

Samarjit Maharana Flow cytometers are very sensitive, there are marked differences in machines even from the same company in assessing fluorescent intensity, and further evidence that the same model from the same company at different locations can have differences. I would say its not easy or correct to simply normalize based on your generated data between the two machines (a quick literature search on flow cytometry quality control will show you this) and when trying to publish your work you may very well get asked about using two different flow cytometers for healthy vs patient population, this is a big red flag, however...

Because of this issue with different machine performance, there are actually protocols to compare performance between machines; see this paper "Methodology for evaluating and comparing flow cytometers: A multisite study of 23 instruments", its open access. The paper will also show you have performance can differ remarkebly even on the same model of flow cytometer, making normlaisation across machines very difficult.

A common approach is to purchase a type of bead called a Antibody capture bead. These beads can capture the antibodies you use, and can be run on different machines to compared performance (typically detection threshold, resolution, etc). I recommend the paper stated above for a more detailed explanation.

DO NOT do what Fatma Abdel Wahab stated; different cytometry machines can have completely different gains, compensation settings, threshold settings, and so on, to reach the same MFI or % expression. Your fluorochromes should be the same however.

If you cannot access the old flow cytometer due to the technical problems, then i recommend re-analysing control samples on the new machine. If you do not have any more healthy controls to re-analyze, consider purchasing healthy control cells, either via blood bank (immune cells), or a cell bank. It may be possible to only re-analyze a few control samples (10 or so) and then attempt data normalisation based on that data. As to how to normalize between old controls and new controls, and then patient samples, there are many statistical approaches. Just make sure you keep the varience (SD, SE, CV) of each group; dont just make the control group a value of 1 with no variance.

David R Parks et al, Interpreting flow cytometry data: a guide for the perplexed