Classification of Systemic Lupus Erythematosus Using Raman Spectroscopy of Blood and Automated Computational Detection Methods: A Novel Tool for Future Diagnostic Testing
Round 1
Reviewer 1 Report
Although I am not a specialist for this type of clinical diagnostics, I have substantial experience from my past work using Raman spectroscopy for analytical taskes applied to biomedical samples, to review this manuscript with reasonable confidence.
Overall, the study seems to rather thourough, and the manuscript is well structured and clearly written. In particular, I find the in-depth description of the methodology quite pointed and clear; subtle points in the analysis and interpretation also seem to be described well, although some are beyond the depth of my understanding of the topic of SLE.
As far as I cane make out there are few typos, which most likely will eb eradiacted by standard spell checking tolls, There is only one typo I identified which might not be caught:
line 159: the pinhole aperure should read um (=micrometer) and not ml (=microliter).
The only other concern I have is related to quite a few of the figures, namely Figures 1 to 5. For many of the graphs the font size of annnotation of the scales and specific features in the graph are next to impossible to read. Of course, on-screen magnification of the open access publication may partially overcome this deficience but then it becomes impossible to view all the figure at once and have the caption in view for consultation about figure details. And many readers prefer to have a printed article in front of them - and then reading these samll fonts is out of the question.
Specifically bad is the situation for the graphs in Figures 1(e), 1(f), 3(a), 4(a)-4(c) and 5(a) - even with a magnifying glass one cannot decipher much of the text.
As far as my experience goes with Figures, most publishers request the minimum font size to be at least 1.5mm or 2mm in print (which is about pt.8 size for most fonts).
Therefore, I REQUEST that the font sizes in Figures 1 to 5 are altered to reasonably comply with this rule of thumb. These changes should be easy to implement, since it is only the text annotation which needs to be treated, not the data plots themselves.
Apart from the improvement in appearance for the figures I think the manuscript content can be published as is.
Author Response
Response to Reviewer 1 Comments
Dear Reviewer,
Thank you for taking the time to review this research manuscript, your suggestions and comments has enabled us to improve and clarify several points within the revised article. These have been described on a point-by-point basis below.
Point 1:
Overall, the study seems to rather thorough, and the manuscript is well structured and clearly written. In particular, I find the in-depth description of the methodology quite pointed and clear; subtle points in the analysis and interpretation also seem to be described well, although some are beyond the depth of my understanding of the topic of SLE. As far as I can make out there are few typos, which most likely will eb eradicated by standard spell checking tolls, There is only one typo I identified which might not be caught:
line 159: the pinhole aperture should read um (=micrometer) and not ml (=microliter).
Response 1:
We have corrected the typographical cited above, thank you for highlighting this to us. A number of spelling and grammatical corrections have been updated in the revised manuscript. We have used the UK spelling for words such as ‘normalisation, standarisation, analyse, hypothesise, aetiological’, however depending on the majority audience we would be happy to update as advised.
Point 2:
The only other concern I have is related to quite a few of the figures, namely Figures 1 to 5. For many of the graphs the font size of annotation of the scales and specific features in the graph are next to impossible to read. Specifically bad is the situation for the graphs in Figures 1(e), 1(f), 3(a), 4(a)-4(c) and 5(a) - even with a magnifying glass one cannot decipher much of the text. As far as my experience goes with Figures, most publishers request the minimum font size to be at least 1.5mm or 2mm in print (which is about pt.8 size for most fonts).
Therefore, I REQUEST that the font sizes in Figures 1 to 5 are altered to reasonably comply with this rule of thumb. These changes should be easy to implement, since it is only the text annotation which needs to be treated, not the data plots themselves. Apart from the improvement in appearance for the figures I think the manuscript content can be published as is.
Response 2:
Thank you for highlighting this to us, we would like all readers to be able to comfortably view the figures and associated annotations. We have reformatted each of the individual data charts, scatterplots and spectral images within Figures 1-5. The advised minimum of size 8 font has been used throughout for figure legends, axis titles and numbering. Further to this we have increased the overall size, marker size and line thickness within some of the charts for clarity, for example in Figure 1 and Figure 4.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This is an interesting paper about Raman-spectroscopy in the diagnosis and even classification of SLE. The technique seems to be easy usable and could advance diagnosis of SLE particularly in unclear cases, since it discriminated well between those having SLE and those that did not (even in the presence of e.g. anti dsDNA antibodies).
Major: It is not clear within the method section, whether the 80 healthy control and 154 validation samples have also been measured spectroscopically? I would have expected so, but I did not find it in the method section and in the text (only in figure 3 is stated : using Raman serum analysis).
And if so, why were these samples not used to model the prediction in figure 4 in order to discriminate between the subgroups?
Furthermore, were these sera all from patients at the timepoint of diagnosis and were multiple longitudinal samples in one patient used to identify disaes and therapy related modifications?
Author Response
Response to Reviewer 2 Comments
Dear Reviewer,
Thank you for taking the time to review this research manuscript, your suggestions and comments has enabled us to improve and clarify several points within the revised article. These have been described on a point-by-point basis below.
Point 1:
It is not clear within the method section, whether the 80 healthy control and 154 validation samples have also been measured spectroscopically? I would have expected so, but I did not find it in the method section and in the text (only in figure 3 is stated : using Raman serum analysis).
Response 1:
We have amended the Materials and Methods section ‘Sample collection and preparation’, ‘Spectral acquisition’ and Table 1 for clarity. Amended Figure 3 legend text in line with this.
In this study, serum samples from eight SLE patients and four healthy controls were collected for spectral analysis. To obtain these measurements, 10 replicate spectra were acquired from duplicate 50ul serum spots, pipetted onto a CaF2 slide giving a total of 20 replicates per participant. The 234 Raman spectra (80 healthy control and 154 SLE spectra) examined within this study were measured using Thermo ScientificTM DXRTM3 dispersive Raman Microscope.
Point 2:
And if so, why were these samples not used to model the prediction in figure 4 in order to discriminate between the subgroups?
Response 2:
Figure 4 is generated using the acquired spectra for the 8 SLE patients, divided into subgroups, and 4 HCs. The total number of spectra collected according to each subgroup has been updated in Table 1. Figure 4 legend text updated for clarity. Previously addressed point 1 should add to improvement of clarity.
Point 3:
Furthermore, were these sera all from patients at the timepoint of diagnosis and were multiple longitudinal samples in one patient used to identify disease and therapy related modifications?
Response 3:
This study used surplus serum samples from patients undergoing routine monitoring tests in our immunology laboratory. We have updated the Materials and Methods section ‘Sample collection and preparation’ for clarity.
Each of these patients had been previously diagnosed with SLE according to classification criteria, therefore samples were not taken at the timepoint of initial diagnosis. Each sample represented a different individual patient. Longitudinal testing was not undertaken as a part of this study
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
All relevant points have been adressed.
