Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The study described in this manuscript is highly relevant. However, there are some inconsistencies that need to be corrected before the manuscript can be considered for publication.
The present study proposes an in silico analysis of the CHIKV E1 gene in isolates from Saudi Arabia, employing phylogenetic analyses, selection analysis, glycosylation prediction, and evaluation of protein stability.
In addition, the authors used a dataset of 109 E1 gene sequences representing the ECSA, Asian, and West African genotypes. For this purpose, phylogenetic analyses using Neighbor-Joining and BEAST were employed, as well as different methods for detecting selective pressure, including MEME, FEL, SLAC, and FUBAR, in addition to structural analysis using DynaMut.

The proposal of the study is interesting. However, in its current form, the manuscript presents some problems and inconsistencies, which need to be adequately revised before eventual acceptance for publication.

 

Major comments

1. Sequence dataset and Table S1

The manuscript presents 109 sequences represented in three genotypes: ECSA (n = 73), Asian (n = 25), and West African (n = 11). However, when evaluating Table S1, I verified an inconsistency with the data presented, since in the table there are 62 sequences for ECSA, 24 sequences for Asian, and 10 sequences for West African.

In addition, I observed the presence of samples from some duplicated genotypes, such as the case of LC259084, which appears related to Tonga (2014) and Philippines (2012), and the accession AF368024, related to S27 (1952) and Tanzania (1953). Both are in the column related to the Asian genotype.

My suggestion is to revise Table S1 and correct the accession number, genotype, and number of sequences obtained, correcting both Table S1 and the information provided in the manuscript.

2. Phylogenetic analysis

For the phylogeny, a Neighbor-Joining tree was constructed with p-distance and bootstrap. This analysis is not ideal to support statements related to sublineage and phylogenetic relationships among the sequences.

In my recommendation, for this type of manuscript it would be more appropriate to use a Maximum Likelihood model. In addition, my recommendation would be to use a more robust software, such as IQ-TREE, RAxML, or FastTree; however, this is only a recommendation.

When analyzing the phylogeny figure, I also noticed that there are sequences with identical numbering, such as AF369024 S27/1952 and AF736024 Tanzania 1953. In addition, it seems to be a possible duplicated sample, since the branch does not show apparent divergence between them.

3. BEAST analysis

The analysis performed in BEAST requires further clarification. Although the authors describe the parameters used, it is not clear how the sampling dates were incorporated into the analysis, whether the temporal signal of the data was evaluated, or whether molecular clock models were compared.

In addition, the value reported as “root height = 0.103 substitutions/site” does not, by itself, represent an estimate of divergence time. Therefore, I recommend a revision of the analyzed data.

4. Results of selective pressure

When analyzing the data present in the manuscript with the data provided in the supplementary material, there appears to be a possible divergence, since the results presented for codon 131 are not evident. In addition, the results presented in the supplementary material could be better presented.

I also observe that, in some methods, such as FEL, information is missing, such as a complete table for a better analysis of the results, because as it currently stands, its interpretation is confusing.

5. Structural part

I noticed that relevant information was missing, such as: was the model made in AlphaFold? What was the quality presented in the prediction of this model? Were preexisting sequences from the PDB used? If so, what is the accession number of this model? Only the GenBank accession number is not sufficient.

6. Data availability

Some extremely important data for the manuscript and its reproducibility were not made available. It would be important for the authors to provide the alignments, the phylogenetic trees, the BEAST .xml file, as well as the logs and associated files.

 

Minor comments

1. The authors need to correct and standardize some terms throughout the manuscript, such as, for example, West African and Western African, since the author uses both terms to refer to the same thing.
2. Also correct the reference to the supplementary table from Tables S1 to Table S1.
3. Verify and correct inconsistencies between the text and the table. For example, in the text, the accession LC259088 is cited for the Tonga (2014) sample, while in the supplementary material table it appears as LC259084.

4. Regarding the figures, more detailed descriptions are missing in the FUBAR analysis figure. I recommend adding a legend explaining the meaning of the size of the dots, the color scale, the highlighted sites, and the threshold used.

 

Final opinion

The manuscript has a relevant question and potential for publication. However, in its current state, I do not recommend its acceptance without first undergoing a series of corrections.

The problems identified in the manuscript go beyond merely textual issues and include inconsistencies in the dataset used, possible duplicated sequences, limitations in the phylogenetic analysis employed, insufficient description of the structural analyses, and absence of fundamental files to ensure the reproducibility of the results.

The authors should carefully revise the dataset, correct the inconsistencies observed in the tables and figures, reassess the phylogenetic analyses, and provide the necessary files to reproduce the results, including alignments, phylogenetic trees, and configuration files used in the analyses.

Author Response

Response to Reviewers’ Comments

Manuscript ID: viruses-4399322

Dear Editor/ Reviewers,

I sincerely thank the Editor and all reviewers for their careful evaluation of the manuscript and for their constructive comments. I have revised the manuscript accordingly and addressed each point in detail below.

Reviewer(s)' Comments to Author:

Reviewer 01

The study described in this manuscript is highly relevant. However, there are some inconsistencies that need to be corrected before the manuscript can be considered for publication. The present study proposes an in silico analysis of the CHIKV E1 gene in isolates from Saudi Arabia, employing phylogenetic analyses, selection analysis, glycosylation prediction, and evaluation of protein stability. In addition, the authors used a dataset of 109 E1 gene sequences representing the ECSA, Asian, and West African genotypes. For this purpose, phylogenetic analyses using Neighbor-Joining and BEAST were employed, as well as different methods for detecting selective pressure, including MEME, FEL, SLAC, and FUBAR, in addition to structural analysis using DynaMut. The proposal of the study is interesting. However, in its current form, the manuscript presents some problems and inconsistencies, which need to be adequately revised before eventual acceptance for publication.

 

 Major comments

• Sequence dataset and Table S1

Comment 01. The manuscript presents 109 sequences represented in three genotypes: ECSA (n = 73), Asian (n = 25), and West African (n = 11). However, when evaluating Table S1, I verified an inconsistency with the data presented, since in the table there are 62 sequences for ECSA, 24 sequences for Asian, and 10 sequences for West African.

Response: I thank the reviewer for their careful attention to the data. I acknowledge the inconsistency and have thoroughly revised Table S1. Upon re‑evaluation of the complete set of 109 sequences, I have corrected the genotype assignments and updated the table accordingly. The new counts are now consistent with the numbers stated in the manuscript: ECSA/IOL = 73 sequences, Asian = 25 sequences and, West African = 11 sequences (Total = 73 + 25 + 11 = 109 sequences, yellow highlighted in section 2.2). I have updated Table S1 to reflect these changes, and the revised table now exactly matches the totals (73, 25, and 11) reported in the main text.

 

Comment 02. In addition, I observed the presence of samples from some duplicated genotypes, such as the case of LC259084, which appears related to Tonga (2014) and Philippines (2012), and the accession AF368024, related to S27 (1952) and Tanzania (1953). Both are in the column related to the Asian genotype. My suggestion is to revise Table S1 and correct the accession number, genotype, and number of sequences obtained, correcting both Table S1 and the information provided in the manuscript.

Response: I thank the reviewer for their meticulous review of the metadata. I have carefully re‑examined each accession number, its associated country/year, and its genotype assignment. The revised Table S1 (attached) has been corrected accordingly, and I confirm that both issues raised by the reviewer are now resolved. I appreciate the reviewer’s valuable input in improving the accuracy of our data.

 

• Phylogenetic analysis

Comment 03. For the phylogeny, a Neighbor-Joining tree was constructed with p-distance and bootstrap. This analysis is not ideal to support statements related to sublineage and phylogenetic relationships among the sequences.

Response: I thank the reviewer for this important methodological recommendation. I have now replaced the Neighbor-Joining analysis with a Maximum Likelihood approach using IQ‑TREE v3.1.3 with 1,000 ultrafast bootstrap replicates. The best‑fit GTR+I model was selected using the Bayesian Information Criterion (BIC = 13,695.527). The new tree provides stronger statistical support for sublineage assignments (100% UFBoot for all major clades) and is now presented as Figure 3. Related parts in methodology (Lines: 150-162), and results (Lines: 295-313) were revised and highlighted in yellow.

 

Comment 04. In my recommendation, for this type of manuscript it would be more appropriate to use a Maximum Likelihood model. In addition, my recommendation would be to use a more robust software, such as IQ-TREE, RAxML, or FastTree; however, this is only a recommendation.

Response: I agree with the reviewer and have implemented the recommended Maximum Likelihood analysis using IQ‑TREE. This approach is now the primary phylogenetic reconstruction in the revised manuscript, with model selection, bootstrap support, and software details clearly stated in the Methods section.

 

Comment 05. When analyzing the phylogeny figure, I also noticed that there are sequences with identical numbering, such as AF369024 S27/1952 and AF736024 Tanzania 1953. In addition, it seems to be a possible duplicated sample, since the branch does not show apparent divergence between them.

Response: I thank the reviewer for this observation. As part of the revised ML analysis (Comment 03), the duplicate branch is therefore no longer present in the updated phylogeny.

• BEAST analysis

Comment 06. The analysis performed in BEAST requires further clarification. Although the authors describe the parameters used, it is not clear how the sampling dates were incorporated into the analysis, whether the temporal signal of the data was evaluated, or whether molecular clock models were compared. In addition, the value reported as “root height = 0.103 substitutions/site” does not, by itself, represent an estimate of divergence time. Therefore, I recommend a revision of the analyzed data.

Response: I thank the reviewer for this constructive comment. To fully address the concerns, I re-ran the BEAST analyses with explicit documentation of all steps. I have thoroughly revised Section 2.5 (Divergence time analysis) as follows:

• Sampling dates: In BEAUti v10.5.0, collection years (1953–2026) were assigned to all 109 taxa via the ‘Tips’ panel, generating <date> tags for every sequence in the XML input files.
• Strict clock and relaxed uncorrelated lognormal clock models were compared through two independent MCMC runs of 10 million generations each.
• I now explicitly distinguish rootHeight (in substitutions/site) from time-calibrated divergence estimates in calendar years, with both values reported in the revised Results (Section 3.3).
• The related sentences in the discussion section were revised and highlighted (Lines: 456-71, highlighted in yellow)

 

 

• Results of selective pressure

Comment 07. When analyzing the data present in the manuscript with the data provided in the supplementary material, there appears to be a possible divergence, since the results presented for codon 131 are not evident. In addition, the results presented in the supplementary material could be better presented. I also observe that, in some methods, such as FEL, information is missing, such as a complete table for a better analysis of the results, because as it currently stands, its interpretation is confusing.

Response: I would like to thank you for your careful evaluation and for identifying the discrepancies between our presented results and the supplementary data. I have thoroughly revised the selection pressure analysis to address all concerns raised (Results section 3.4 and discussion Lines: 455-480; highlighted in yellow ). The following corrections and improvements were applied:

• Corrected codon 131 error: After re-examining the raw MEME output, I confirm that codon 131 (previously reported as p = 0.048) was incorrectly identified. The correct MEME-significant site iscodon 99 (p = 0.01, β⁺ = 134.01 on ~16% of branches). This error has been corrected throughout the manuscript.
• Improved supplementary data presentation: I have now arranged all supplementary data into dedicated, clearly labeled folders that directly correspond to the specific sections of the manuscript. The current organization is as follows:

• Section 2.2. - Table S1 – Contains the detailed sequence information and metadata used for the alignment.
• Section 2.7. (DynMut, DUET and mCSM Analysis) – Houses the complete output files for all structural stability predictions, including DynaMut stability scores, DUET consensus predictions, and mCSM mutation effect summaries, along with the corresponding PDB and session files for each variant.
• Section 3.3. (BEAST Analysis) – Contains the Newick tree files generated from both the strict clock and relaxed uncorrelated lognormal clock models used for divergence time estimation.
• Section 3.4 (FUBAR, FEL, MEME, SLAC) – Includes the complete raw output tables and result files from all four codon-based selection pressure detection methods.
• Section 3.4 - Figures S6 – Provides FUBAR Plot as requested in comment 15.
• Section 3.5 - Figures S1-S5 – Results for six additional E1 mutations (K211E, K211N, M269V, D284E, I317V and V322A) which include their predicted ΔΔG values and interatomic interaction profiles.

 

• Added complete FEL table: Previously missing, the full FEL results table is now included in the supplementary materials with all relevant statistical metrics (α, β, LRT, p-value, and classification).
• Clarified interpretation: The revised text now provides clearer interpretation of each method's output, including:
• A summary table comparing method results.
• Clear categorization of purifying (125 sites in FEL, 140 in FUBAR, 58 in SLAC) and positive selection sites.
• Explicit statement that no FEL-significant positive sites were detected at p < 0.05.

 

Again I appreciate the reviewer’s valuable input in improving the accuracy of our data.

 

• Structural part

Comment 10. I noticed that relevant information was missing, such as: was the model made in AlphaFold? What was the quality presented in the prediction of this model? Were preexisting sequences from the PDB used? If so, what is the accession number of this model? Only the GenBank accession number is not sufficient.

Response: Thank you for this important comment. In response, I have revised Section 2.7 to provide complete transparency regarding the structural modeling approach. A three-dimensional structural model of the wild-type E1 protein was generated via comparative (homology) modeling using the SWISS-MODEL server based on the prototype strain S27 sequence (GenBank accession AF485728). Model quality was assessed using the Global Model Quality Estimation (GMQE = 0.88) and QMEANDisCo Global score (0.83 ± 0.05), with stereochemical quality validated by Ramachandran plot analysis. Regarding the use of preexisting PDB structures, SWISS-MODER uses PDB templates for homology modeling; the template used was the cryo-EM structure of the CHIKV E1 glycoprotein, which shares 99.22% sequence identity with the target S27 sequence. The GenBank accession number (AF485728) for the target sequence has been retained, and the PDB template identifier has been added to provide complete accession information. These additions ensure that the model construction, quality metrics, and template information are now fully documented in the revised manuscript.

 

• Data availability

Comment 11. Some extremely important data for the manuscript and its reproducibility were not made available. It would be important for the authors to provide the alignments, the phylogenetic trees, the BEAST .xml file, as well as the logs and associated files.

Response: I agree that these files are essential for reproducibility. All requested materials are now provided in the supplementary data:

• Alignments → Sequence alignments folder (FASTA and NEXUS formats).
• Phylogenetic trees (Newick) → FigTree phylogram Newick Files folder contains both the ML tree and BEAST MCC trees (strict and relaxed) as Newick‑formatted files with branch lengths and node supports.
• BEAST XML, logs, and associated files (trees, state logs, operator logs) → fully deposited in the BEAST Analysis folder.
• Selection pressure outputs → Selection pressure analysis folder.
• Stability predictions → DynMut, DUET and mCSM Analysis folder.

All are now cited in the revised Data Availability section and supplementary legends.

 

 

Minor comments

Comment 12. The authors need to correct and standardize some terms throughout the manuscript, such as, for example, West African and Western African, since the author uses both terms to refer to the same thing.

Response: I thank the reviewer for this observation. I have standardized the terminology throughout the manuscript by replacing all instances of "Western African" with "West African" to ensure consistency with both scientific convention and established genotype nomenclature. 

Comment 13. Also correct the reference to the supplementary table from Tables S1 to Table S1.

Response: Thank you for this observation. I have corrected the reference to the supplementary table from "Tables S1" to "Table S1" throughout the manuscript to ensure proper singular/plural agreement.

Comment 14. Verify and correct inconsistencies between the text and the table. For example, in the text, the accession LC259088 is cited for the Tonga (2014) sample, while in the supplementary material table it appears as LC259084.

Response: Thank you for your careful review and for identifying the accession number discrepancy. I have thoroughly re‑examined all accession numbers, their associated countries, years, and genotype assignments. The specific inconsistency between the text (LC259088) and the supplementary table (LC259084) for the Tonga (2014) sample has been corrected. I appreciate the reviewer's valuable input in improving the accuracy of our data.

 

Comment 15. Regarding the figures, more detailed descriptions are missing in the FUBAR analysis figure. I recommend adding a legend explaining the meaning of the size of the dots, the color scale, the highlighted sites, and the threshold used.

Response: I sincerely thank you for this valuable suggestion. The FUBAR analysis figure is already included in the supplementary materials as Figure S6 with legend explanation and also cited in the text in section 3.4.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

In “Molecular evolution of the Chikungunya virus E1 gene in Saudi Arabia: predominance of purifying selection and ECSA/IOL lineage circulation,” the author analyzes publicly available CHIKV E1 sequences, including seven Saudi Arabian isolates, to infer genotype placement, phylogenetic relationships, selection pressure, glycosylation motifs, and predicted stability effects of selected E1 mutations. The study addresses a relevant regional surveillance gap and the main finding that available Saudi CHIKV E1 sequences cluster within the ECSA/IOL lineage is clear. Overall, the manuscript mainly requires clearer methods and improved figure presentation.

Comments

• Methods lines 119–130: The Methods section should clarify the sequence selection criteria a little more explicitly, especially how duplicates, short fragments, and ambiguous sequences were defined and filtered 
• Figure 2 is informative but crowded; the visual emphasis on the Saudi-unique substitutions and N-glycosylation site could be made clearer
• Table 1 would benefit from a slightly clearer legend defining the abbreviations and the meaning of the mixed outputs across MEME, FEL, SLAC, and FUBAR, especially for readers who do not use HyPhy regularly

Author Response

Response to Reviewers’ Comments

Manuscript ID: viruses-4399322

Dear Editor/ Reviewers,

I sincerely thank the Editor and all reviewers for their careful evaluation of the manuscript and for their constructive comments. I have revised the manuscript accordingly and addressed each point in detail below.

Reviewer(s)' Comments to Author:

Reviewer 02

In “Molecular evolution of the Chikungunya virus E1 gene in Saudi Arabia: predominance of purifying selection and ECSA/IOL lineage circulation,” the author analyzes publicly available CHIKV E1 sequences, including seven Saudi Arabian isolates, to infer genotype placement, phylogenetic relationships, selection pressure, glycosylation motifs, and predicted stability effects of selected E1 mutations. The study addresses a relevant regional surveillance gap and the main finding that available Saudi CHIKV E1 sequences cluster within the ECSA/IOL lineage is clear. Overall, the manuscript mainly requires clearer methods and improved figure presentation.

 

Comments

Comment 01. Methods lines 119–130: The Methods section should clarify the sequence selection criteria a little more explicitly, especially how duplicates, short fragments, and ambiguous sequences were defined and filtered.

Response: I thank you for this important comment. In response, I have revised Section 2.2 to explicitly define our sequence filtering criteria: duplicates were removed based on 100% nucleotide identity and identical metadata; short fragments were excluded if shorter than the target 1155 bp analysis region; and sequences with ≥ 1% ambiguous bases (Ns) were discarded.  

 

Comment 02. Figure 2 is informative but crowded; the visual emphasis on the Saudi-unique substitutions and N-glycosylation site could be made clearer

Response: Thank you for this valuable suggestion. In response, I have revised Figure 2 by removing some sequences from each representative genotype block to reduce crowding and allow for clearer visualization. This revision enhances the visual emphasis on the Saudi-unique substitutions and N-glycosylation sites, making these key features more distinct and interpretable.

 

Comment 03. Table 1 would benefit from a slightly clearer legend defining the abbreviations and the meaning of the mixed outputs across MEME, FEL, SLAC, and FUBAR, especially for readers who do not use HyPhy regularly.

Response: Thank you for this constructive suggestion. In response, I have revised Table 1 with an expanded legend that clearly defines all abbreviations (MEME, FEL, SLAC, FUBAR) and explains the interpretation of mixed outputs across methods.

 

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The new version of the manuscript is substantially improved compared with the previous version. The authors made all the corrections previously requested and adequately addressed the vast majority of the comments, significantly improving the overall quality of the work. However, some additional points remain, as presented below:

1. The authors improved and incorporated the sampling dates and performed both strict and relaxed clock models. However, I suggest that the authors present an appropriate assessment, such as root-to-tip regression and a date-randomization analysis. If sufficient temporal signal is not observed, I recommend removing these estimates and retaining only the non-temporal phylogenetic reconstruction.

2. There are inconsistencies between the values reported in the text and those presented in Table 1. The text states that codon 195 had a posterior probability (p.p.) of 0.942 according to FUBAR, whereas Table 1 reports a p.p. value of 0.022. In addition, the results reported for codons 298 and 332 also differ from those presented in Table 1. Therefore, I suggest that the authors carefully review the reported data.

3. The authors suggest multiple routes of introduction of the viral strains into Saudi Arabia. This hypothesis is plausible; however, with only seven Saudi sequences and using only a fragment of the E1 gene, it is not possible to make this claim, since the findings may also result from local circulation. Therefore, I recommend that the authors revise this statement.

4. The authors use a 960-bp alignment for the maximum-likelihood analysis, whereas they use 1,155 bp for the BEAST analysis. This difference needs to be justified, particularly because the two trees are compared in the interpretation of the results.

5. I recommend checking the versions of the BEAST, BEAUti, and TreeAnnotator software. The version “v10.5.0” is incorrect.

6. In Figure 4, the legend states that branch lengths represent time in years, whereas the scale bar represents substitutions per site. I suggest providing a clearer explanation.

Author Response

Response to Reviewers’ Comments

Manuscript ID: viruses-4399322

Dear Editor/ Reviewers,

I sincerely thank the Editor and all reviewers for their careful evaluation of the manuscript and for their constructive comments. I have revised the manuscript accordingly and addressed each point in detail below.

Reviewer(s)' Comments to Author:

Reviewer 01

 The new version of the manuscript is substantially improved compared with the previous version. The authors made all the corrections previously requested and adequately addressed the vast majority of the comments, significantly improving the overall quality of the work. However, some additional points remain, as presented below:

Comment 01. The authors improved and incorporated the sampling dates and performed both strict and relaxed clock models. However, I suggest that the authors present an appropriate assessment, such as root-to-tip regression and a date-randomization analysis. If sufficient temporal signal is not observed, I recommend removing these estimates and retaining only the non-temporal phylogenetic reconstruction.

Response: I sincerely thank the reviewer for this critical suggestion. Following your recommendation, I have performed a root-to-tip regression analysis using TempEst v1.5.3 to formally assess the temporal signal in our E1 dataset. The analysis revealed a very weak negative correlation between genetic distance and sampling time (R² = 1.26 × 10⁻³; correlation coefficient = −0.0355), indicating a complete absence of usable temporal signal for reliable molecular clock calibration. This analysis is now described in (Section 2.5. Assessment of Root-to-Tip Regression and Temporal Signal). Accordingly, I have removed all time-calibrated divergence estimates (strict and relaxed clock analyses) from the manuscript. I retain only the non-temporal phylogenetic reconstructions (maximum likelihood tree, Figure 3), which are robust and fully support our conclusions regarding genotype assignment and lineage circulation. All mentions of specific divergence dates (e.g., root age, tMRCA of Saudi strains) have been deleted from the Abstract, Results, and Discussion. I agree that this cautious approach prevents overinterpretation and strengthens the scientific integrity of the study.

 

Table. TempEst v1.5.3 temporal signal result

Parameter	Value
Date range	74
Slope (rate)	-0
X-Intercept (TMR...)	200.375
Correlation Coefficient	-0.0355
R squared	1.2595E-3
Residual Mean Sq...	2.2508E-4

 

Comment 02. There are inconsistencies between the values reported in the text and those presented in Table 1. The text states that codon 195 had a posterior probability (p.p.) of 0.942 according to FUBAR, whereas Table 1 reports a p.p. value of 0.022. In addition, the results reported for codons 298 and 332 also differ from those presented in Table 1. Therefore, I suggest that the authors carefully review the reported data.

Response: I thank the reviewer for carefully identifying this inconsistency. Upon thorough review of our raw output files (FEL Table.docx, FUBAR.csv, MEME.csv, SLAC.csv), I have corrected Table 1 and the corresponding text in Section 3.3, Highlighted in the MS. All other values in Table 1 have been verified against the raw output files and are corrected.  I apologize for this oversight and thank the reviewer for their attention to detail.

 

Comment 03. The authors suggest multiple routes of introduction of the viral strains into Saudi Arabia. This hypothesis is plausible; however, with only seven Saudi sequences and using only a fragment of the E1 gene, it is not possible to make this claim, since the findings may also result from local circulation. Therefore, I recommend that the authors revise this statement.

Response: I thank the reviewer for this important and valid critique. I agree that with only seven Saudi sequences from a single city (Jeddah), covering a limited time frame (2018–2021), and using only the E1 gene fragment, definitive conclusions about multiple introductions versus local circulation are not warranted.

Accordingly, I have revised the language throughout the manuscript to present the "multiple introductions" scenario as a plausible interpretation rather than a concluded fact. Specific changes included the Abstract (lines: 20-23, highlighted in yellow), Results (Section 3.2, Lines 294-296, highlighted in yellow), Discussion (lines: 412-416 and lines 512-518, highlighted in yellow), and the conclusion (lines: 539-542 highlighted in yellow ).

I am very grateful to the reviewer for prompting this important qualification, which strengthens the scientific rigor of the study.

Comment 04. The authors use a 960-bp alignment for the maximum-likelihood analysis, whereas they use 1,155 bp for the BEAST analysis. This difference needs to be justified, particularly because the two trees are compared in the interpretation of the results.

5. I recommend checking the versions of the BEAST, BEAUti, and TreeAnnotator software. The version “v10.5.0” is incorrect.

Response: I thank the reviewer for this observation. The 1,155 bp fragment represented the full E1 coding sequence, while the 960 bp alignment used for phylogenetic analysis excluded the hypervariable signal peptide (first 150 bp) and transmembrane domain (last 45 bp) to reduce noise and improve alignment accuracy. However, since our temporal signal assessment revealed insufficient clock-like structure I have removed all BEAST molecular clock analyses from the revised manuscript. Consequently, only the 960 bp alignment is now used for the final phylogenetic reconstruction (IQ‑TREE), eliminating the discrepancy between alignment lengths.

Comment 05. In Figure 4, the legend states that branch lengths represent time in years, whereas the scale bar represents substitutions per site. I suggest providing a clearer explanation.

Response: I thank the reviewer for catching this inconsistency. As noted in our response to Comment 01, I have removed Figure 4 (time-calibrated Bayesian tree) from the revised manuscript because temporal signal was insufficient for reliable molecular clock dating. Therefore, this figure and its legend are no longer included. Only the maximum likelihood tree (Figure 3) remains.

Author Response File: Author Response.pdf

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

The new version of the manuscript has improved compared with the previous version, and most of the comments were adequately addressed. However, some additional points remain, as presented below:

 

1. There are still some inconsistencies in the text that need to be corrected. In Section 3.3, lines 328–340, and in Table 1, codon 195 is classified as being under positive selection by FUBAR, although the reported posterior probability is 0.022, which is below the minimum threshold of 0.9 defined in lines 183–186. In the Discussion, lines 429–434, different values are presented for codons 195, 298, and 332. The authors should correct the values and interpretations in Section 3.3, Table 1, and the Discussion. If the values are not supported by at least two methods, I suggest removing or reinterpreting these results.

 

2. Some statements regarding multiple introductions still need to be corrected. Although lines 412–416 acknowledge the limitations of the dataset, lines 421–420 state that the phylogenetic pattern “clearly support repeated, relatively recent introductions rather than sustained long-term endemic circulation”. This sentence contradicts the previous caveat. I recommend revising this passage and other similar statements in the Discussion and Conclusion.

 

3. The figure numbering should be corrected. Figure 5 should be renumbered as Figure 4. Likewise, all citations in the text should also be corrected.

Author Response

Response to Reviewers’ Comments

Manuscript ID: viruses-4399322

Dear Editor/ Reviewers,

I sincerely thank the Editor and all reviewers for their careful evaluation of the manuscript and for their constructive comments. I have revised the manuscript accordingly and addressed each point in detail below.

Reviewer(s)' Comments to Author:

Reviewer 01

The new version of the manuscript has improved compared with the previous version, and most of the comments were adequately addressed. However, some additional points remain, as presented below:

 

Comment 01. There are still some inconsistencies in the text that need to be corrected. In Section 3.3, lines 328–340, and in Table 1, codon 195 is classified as being under positive selection by FUBAR, although the reported posterior probability is 0.022, which is below the minimum threshold of 0.9 defined in lines 183–186. In the Discussion, lines 429–434, different values are presented for codons 195, 298, and 332. The authors should correct the values and interpretations in Section 3.3, Table 1, and the Discussion. If the values are not supported by at least two methods, I suggest removing or reinterpreting these results.

 

Response: I would like to thank the reviewer for the precise and accurate revision and for his patience reporting and catching such inconsistencies. I have thoroughly revised all relevant sections to ensure strict consistency with my predefined criterion (positive selection requires p < 0.1 in MEME, FEL, or SLAC, or FUBAR posterior probability > 0.9, and corroboration by at least two methods).

Specifically:

1. Section 3.3: I have corrected the interpretation of codon 195 from "pervasive positive selection" to purifying selection, as its FUBAR positive posterior probability is 0.022 (well below the 0.9 threshold) and SLAC P[dN/dS < 1] = 1.000. I have also reclassified codons 298 and 332 as purifying rather than "moderate positive candidates," as they fail the ≥2 method corroboration requirement. MEME-only signals at codons 99, 306, and 334 are now explicitly described as isolated, single‑method observations requiring cautious interpretation.
2. Table 1: The "Interpretation" column has been updated accordingly. Codon 195 is now labeled as "Purifying", with the supporting evidence (SLAC P[dN/dS < 1] = 1.000, FUBAR positive PP = 0.022) clearly reflected. Codons 298 and 332 are also labeled as "Purifying". Codons 99, 306, and 334 are labeled as "MEME‑only signals" to indicate they fail the ≥2 method rule. A footnote has been added to the table explicitly stating that no codon met the threshold for robustly supported positive selection.
3. Discussion: I have corrected the values and interpretations for codons 195, 298, and 332 to match Table 1 and Section 3.3. The incorrect claim of positive selection at codon 195 has been removed, and the language regarding codons 298 and 332 has been revised to reflect their classification as purifying sites.

Comment 02. Some statements regarding multiple introductions still need to be corrected. Although lines 412–416 acknowledge the limitations of the dataset, lines 421–420 state that the phylogenetic pattern “clearly support repeated, relatively recent introductions rather than sustained long-term endemic circulation”. This sentence contradicts the previous caveat. I recommend revising this passage and other similar statements in the Discussion and Conclusion.

 

Response: I thank the reviewer for this careful observation. I have revised the relevant passage in the Discussion to eliminate the overstatement. Specifically, I have replaced the phrase "clearly support" with "are consistent with" and integrated the limitation directly into the same sentence ("although sustained local transmission cannot be definitively excluded given the limited sampling"). I have also reviewed the Abstract, Results, and Conclusion to ensure all statements regarding multiple introductions are appropriately cautious and consistent with the acknowledged limitations of the dataset. All changes are highlighted in the revised manuscript.

 

Comment 03. The figure numbering should be corrected. Figure 5 should be renumbered as Figure 4. Likewise, all citations in the text should also be corrected.

Response: I thank the reviewer for this careful observation. I have corrected the figure numbering throughout the manuscript: Figure 5 has been renumbered as Figure 4, and all corresponding in-text citations have been updated accordingly.

Author Response File: Author Response.pdf