The manuscript from Robichon and colleagues, which describes the development of a 24-color Ab panel to be used with 3 laser (BRV) Aurora analyzers tries to build a tool to address important questions in the multiple sclerosis field by inspecting the levels at which many leukocyte subsets express surface markers that are targets of MS drugs (CD49d, S1PRs, CD20)
Nevertheless, I think the paper has several issues, mostly methodological, which makes the panel useless and preclude publication in a journal such as Cytometry Part A, for which methodological soundness is chiefly important

Points related to the methodology (see the pdf file included in the review form)

During review of the .fcs file submitted to flowrepository, (a single FCS file already unmixed, which makes the review difficult) I noticed fluorescent signals of suspect dye aggregates as “super positive” events. An adjunct step of Ab-aggregate clearing by high-speed centrifugation is warranted for flow panels involving many markers.

Complete sets of raw and unmixed files should be deposited in flowrepository. The sample deposited does not contain any signal in the S1PR1 and 4

The unmixing has many errors. Too many for a paper submitted to Cytometry Part A. For instance, I noticed “compensation” errors between CD16 and many fluorochromes of the Violet laser. Is this due to the choice of using the low level CD16 signal from NK cells, instead of the brighter signal from neutrophils, as positive for unmixing? Amid other issues, the unmixing failure resulted also in the odd CD3 ALEXA700 VS CD8 APC-H7 plot, also visible in the last plot of the supplementary figure 6. Importantly, other errors in the unmixing have consequences on interpretation of the data. For instance, T cells have the same CD86 levels of monocytes, which is biologically inaccurate, to say the least. B cells too have more CD86 than monocytes?

CD86 BUV737 is used in the absence of a UV laser.

To evaluate the ability to distinguish positive from negative events in key markers within polychromatic panels, FMO controls should be employed. The comparisons made in the manuscript between single stains and complete panel is thus not sufficient, as they may underestimate the global contribution of all the other fluorochromes to the spreading error of the negative

Titrations
-is the 2 ul titer taken from a different donor for CD3 AF700? Also, in the CD8 APC-H7 titration, the 5 ul titer has less cd8+ t cells compare to the others. All in all, it seems that authors used different donors, and may be in different acquisition session, to be concatenated in the titration plots. I strongly suggest to perform titration of a given antibody using the same donor
-A strange pattern of negative populations is seen in the ccr7 bv421 titration
-Include CD127 in the backbone with CD4 to visualize CD25 vs CD127 in CD4+ t cells. It can enhance understanding on why the 2.5 ul dilution is chosen for CD25.

It’s not useful to make comparisons between 3 HD and 3 MS, since the number of variables analyzed is very large. No p-value adjustment for multiple comparisons is made throughout the study, which is necessary to exclude type 1 statistical errors. Moreover, a paired T test is used for all the comparisons but age (unpaired). Why? What’s the rationale for doing a paired T test for independent samples?

Supplementary figure 9I and 10Z have no relevance and are at least incorrect, as they pool together different variables. Did the authors consult a statistician? Furthermore, the first sentence of supplementary figure 9, 10, and 11 is the same and is incorrectly worded

Line 76 The mention of the last author study on clozapine (ref 8) is out of scope for the submitted paper, as no dopamine nor serotonin receptor expression is evaluated by the present Ab panel
Line 87 May be "crucial", not “curial”
Line 198 “expression patterns”? Better “density levels”?

Figure 1
Ab-secreting cells, can also be CD27- (i.e. prePB)
for cd20+ gated b cells, axes or populations are mislabeled. Also, in case axes labels and not gates are correct (i.e. x=CD38, y=CD27), as I can understand, the CD38 signal seems not to be sufficient to gate for Transitional (CD38hi CD27-) B cells
In the gating of T cell subsets there are many undefined populations. I suggest to name also single-positive cells in the CD62L vs CCR7 plot of gated cd45RA- non-treg. The same, and to a larger extent given the high frequency of cd62l single-positives, applies to cd45ra cd8 t cells
Figure 2
Panel A. there is a near complete absence of s1pr1 expression in healthy donor cd19+ b cells, which is somewhat strange.

Supplementary Figure 2. There is discrepancy between what described in the manuscript and what is depicted in the figure. row 1: cd80 or cd86?

Supplementary Figure 8. No X axes labels are indicated

Lines 306-308. Immunological Conclusions should not be drawn with this small cohort

Table 4. CCR2 and HLA-DR are reported twice