Metagenome Techniques for Detection of Pathogens Causing Ocular Infection

Round 1

Reviewer 1 Report

The prompt and precise diagnosis for ocular infection is crucial for timely treatment and optimized outcomes. However, the current gold standard of culture, stain and PCR for pathogen detection is time consuming and empirical. In this manuscript, the authors compared the diagnostic performance of conventional microbial examinations and shotgun metagenomic analysis (SMA) using ocular specimens, and endorsed the potential of SMA to facilitate the comprehensive detection of causative pathogens in ocular infection. This study is a timely piece of work with substantial clinical importance. Overall the manuscript is well written and data properly interpreted. I have only the following minor points that the authors can consider commenting on or discussing in more details.  

1. An adequate number of reads is essential for precise alignment and thus meaningful results. What’s the reason for huge variations in the number of SMA reads among samples? What is the detection cutoff threshold?

2. In case 7 and 12, Propionibacterium spp. was excluded from causative pathogen due to its commensal nature. However, P. acnes has been known as a frequent cause of bacterial keratitis, and may even present in mixed bacterial and fungal microbial keratitis. How to identify the real causative agent(s) when multiple pathogens are detected?

3. It would be interesting if the authors can discuss how to balance the sensitivity and specificity of pathogen identification while interpreting SMA results.

4. Numerous typos:

L93: Reports of “rentat”

L124: Case 16 patient age (37) is not consistent with that in Table 1 (38)

L119; L176: “Propionibacterium” spp. , : The genus should be italicized in compliance with Bacteriological Code

L167, L172: “probable” pathogen

L189: “there is no need to do not consider”: “do not” should be crossed out.

L216: “Thus Paracoccus spp. ‘may because of’ intraocular inflammation”: should be ‘may be the cause of’

Author Response

Comment: The prompt and precise diagnosis for ocular infection is crucial for timely treatment and optimized outcomes. However, the current gold standard of culture, stain and PCR for pathogen detection is time consuming and empirical. In this manuscript, the authors compared the diagnostic performance of conventional microbial examinations and shotgun metagenomic analysis (SMA) using ocular specimens, and endorsed the potential of SMA to facilitate the comprehensive detection of causative pathogens in ocular infection. This study is a timely piece of work with substantial clinical importance. Overall the manuscript is well written and data properly interpreted. 

Response: We wish to thank the reviewer for the kind evaluation of our study.

Comment 1: An adequate number of reads is essential for precise alignment and thus meaningful results. What’s the reason for huge variations in the number of SMA reads among samples? What is the detection cutoff threshold?

Response: I agree with your suggestion. We used clinical samples stocked at fridge, and size of samples such as corneal scraping might be various. Moreover, severity of infection also was not similar. Thus, samples might include huge variations in the number of SMA reads. We have not examined SMA against control sample (healthy corneal scrapings), and it is very difficult to set the detection cutoff threshold. In this study, all data acquired　by SMA was used for analysis. We added limitations in discussion. (Line 266-268)  

Comment 2: In case 7 and 12, Propionibacterium spp. was excluded from causative pathogen due to its commensal nature. However, P. acnes has been known as a frequent cause of bacterial keratitis, and may even present in mixed bacterial and fungal microbial keratitis. How to identify the real causative agent(s) when multiple pathogens are detected?

Response:　I agree that P. acnes may occur sometimes infection at ocular site. However, exact pathogenesis of ocular infection caused by P. acnes is not understood. Almost samples included reads of P. acnes, and It is very difficult to distinguish between pathogenic and commensal P. acnes. Since we would like to analyze easily SMA data, we excluded P. acnes from causative pathogen. Further investigation about identification of the real causative agent in SMA should be needed. We added limitations in discussion. (Line 240-247)

Comment 3: It would be interesting if the authors can discuss how to balance the sensitivity and specificity of pathogen identification while interpreting SMA results.

Response: Thank you for your suggestion. I added the discussion about balance the sensitivity and specificity of pathogen identification. (Line 275-278)

Comment 4: Numerous typos:

Response:  I corrected typos.

Reviewer 2 Report

The article entitled « Metagenome techniques for detection of pathogens causing ocular infection » reports the investigation of the interest of metagenomic analysis for the diagnosis of ocular infections in comparison with classical techniques.

The description of the clinical cases, treatment, outcomes and of course diagnosis are well described and interesting.

I have just some minor remarks.

The references should be indicated before the dot at the end of the sentence and not at the beginning of the next sentence (cf line 39, line 51, line 60, line 74, the authors should check this point through the document).

In tables 1 and 2, the line width should be the same for all the lines and tables should appear complete in a page.

In figure 2 legend , please modify « enkaryotic » by eukaryotic

Line 166 to 169, the sentence is too long and should be reformulated

Line 247, « mix » should be modified by mixed

Line 271 to 281 : the dates of the prospective study should be mentionned

Author Response

Comment: The article entitled « Metagenome techniques for detection of pathogens causing ocular infection » reports the investigation of the interest of metagenomic analysis for the diagnosis of ocular infections in comparison with classical techniques. The description of the clinical cases, treatment, outcomes and of course diagnosis are well described and interesting.

Response: We wish to thank the reviewer for the kind evaluation of our study.

Comment 1: The references should be indicated before the dot at the end of the sentence and not at the beginning of the next sentence (cf line 39, line 51, line 60, line 74, the authors should check this point through the document).

Response: I changed as suggested.

Comment 2: In tables 1 and 2, the line width should be the same for all the lines and tables should appear complete in a page.

Response: I changed as suggested.

Comment 3: In figure 2 legend , please modify « enkaryotic » by eukaryotic

Response: I corrected typo.

Comment 4: Line 166 to 169, the sentence is too long and should be reformulated

Response: I reformulated as suggested. (Line 166-170)

Comment 5: Line 247, « mix » should be modified by mixed

Response: I changed as suggested.

Comment 6:Line 271 to 281 : the dates of the prospective study should be mentioned

Response: I added dates of the study. (Line 294)

Reviewer 3 Report

In the current case report "Metagenome techniques for detection of pathogens causing ocular infection" the authors evaluated the use of metagenomics techniques to diagnose the ocular infection. The authors compared the metagenomics diagnosis with the conventional diagnostic methods. The authors showed the higher potential of metagenomics in ocular infection. The manuscript is of clinical significance, but have the following queries to use of metagenomics in ocular diagnostics.

1) The ocular infection such as keratitis and endophthalmitis need rapid diagnosis and prompt treatment to save vision loss. The culture methods need time for diagnosis and often difficult for the culture-negative cases, therefore the PCR technique is rapid and does not need advanced machines and bioinformatics analysis experts. How authors justify the use of metagenomics for ocular diagnosis as it is time-consuming, costly, and needs an expert in experimental as well data analysis? Authors should also elaborate on these in the manuscript.

2) Rather than paired-end sequencing why authors preferred single-end sequencing? As in the discussion, the authors mentioned that they encountered the unclassified reads in the study that could have averted by the paired-end reads.

3) Authors mentioned in line 265 "cut-off value of reads must be set to determine pathogenic or non-pathogenic agent". How more or less reads can determine the pathogenicity?

3) Authors should use italics for the microbial species throughout the manuscript (eg. line 170 and 171).

 

Author Response

Comment 1: The ocular infection such as keratitis and endophthalmitis need rapid diagnosis and prompt treatment to save vision loss. The culture methods need time for diagnosis and often difficult for the culture-negative cases, therefore the PCR technique is rapid and does not need advanced machines and bioinformatics analysis experts. How authors justify the use of metagenomics for ocular diagnosis as it is time-consuming, costly, and needs an expert in experimental as well data analysis? Authors should also elaborate on these in the manuscript.

Response: Thank you for your suggestion. I added advantage and disadvantage of metagenomics for ocular diagnosis in discussion. (Line 282-286)

Comment 2: Rather than paired-end sequencing why authors preferred single-end sequencing? As in the discussion, the authors mentioned that they encountered the unclassified reads in the study that could have averted by the paired-end reads.

Response: Thank you for your question. We analyzed single-end sequencing because of cost reduction. Moreover, unclassified reads generally include a lot of repeated sequences and paired-end sequencing may not reduce unclassified reads. I added above discussion. (Line 234-235)

Comment 3: Authors mentioned in line 265 "cut-off value of reads must be set to determine pathogenic or non-pathogenic agent". How more or less reads can determine the pathogenicity?

Response: We have not examined SMA against control sample (healthy corneal scrapings), and it is very difficult to set the detection cutoff threshold. Further investigation about determination of pathogenic or non-pathogenic agent should be needed. We added limitations in discussion. (Line 271-278)

Comment 4: Authors should use italics for the microbial species throughout the manuscript (eg. line 170 and 171).

Response: I changed as suggested.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.