Viruses, Vectors, and Villains: Governing the Risks and Rewards of Artificial Intelligence in Virology

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors

1. The background section talks about many AI tools and technologies, which is useful but a bit hard to follow in text alone. A simple figure that visually groups these tools (for example, by their use in diagnosis, surveillance, drug discovery, and synthetic biology) would make the ideas much easier for readers to understand.
2. When discussing how policies should be designed for AI-assisted virology, a clear flowchart showing how AI is used, where risks arise, and where oversight or decision-making happens would help readers quickly grasp how the proposed governance framework would work in practice.

Author Response

Comment 1.1: "The background section talks about many AI tools and technologies, which is useful but a bit hard to follow in text alone. A simple figure that visually groups these tools (for example, by their use in diagnosis, surveillance, drug discovery, and synthetic biology) would make the ideas much easier for readers to understand."

Response: We agree and have added Figure 1 (line 52) titled "The roles of AI in modern virology." This figure visually groups AI tools by their major application areas including diagnostics, surveillance, drug discovery, and synthetic biology. The figure is referenced in the text and includes a descriptive caption explaining how AI is being deployed across these domains.

Comment 1.2: "When discussing how policies should be designed for AI-assisted virology, a clear flowchart showing how AI is used, where risks arise, and where oversight or decision-making happens would help readers quickly grasp how the proposed governance framework would work in practice."

Response: Excellent suggestion. We have added Figure 2 titled "Proposed governance framework for AI-assisted virology." This flowchart illustrates the relationship between international bodies (UN, WHO), multidisciplinary advisory panels, policymakers, and primary users (researchers, hospitals). The figure shows how policies are developed, standards are updated, incentives are provided, and how AI systems are applied while minimizing risks. The figure is referenced in the text and includes a detailed caption explaining the governance workflow.

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript entitled “Viruses, Vectors, and Villains: Governing the Risks and Rewards of Artificial Intelligence in Virology” the authors discuss the need for ethical AI standards in the health domain while addressing the dual-use risks of AI in virology. The authors suggest the how can health governance bodies be empowered to develop and manage automated data pipelines and open-access AI tools. The authors identify critical gaps in current governance frameworks and propose a comprehensive, UN-led approach. The proposed framework includes (a) an integrated AI-assisted surveillance network for outbreak attribution, (b) standardized screening protocols for DNA synthesis providers, (c) preclinical validation standards for AI-frameworks, and (d) international programs incentivizing responsible practice. The authors suggest that this framework aims to balance scientific innovation with biosecurity. The authors also acknowledge fundamental challenges such as, misaligned economic incentives for voluntary compliance, uneven implementation capacity across nations, and the need to balance scientific freedom with biosecurity. Overall, the review is well organized, easy to follow, and well motivated. The manuscript can be further improved if the authors could discuss and address the following suggestions and comments. 

1. Authors mention: “A panel of experts should be formed to manually validate sequences flagged as synthetically altered, and output formats should be graphical and easily understood.” -- manually validating the sequences may also lead to human errors and  higher expenses, inclusion of cost-benefit analysis of proposed framework would highly support the authors views. 
2. Not all AI-generated sequences, structures, or compounds can be readily synthesized and experimentally validated. Additionally, a significant knowledge gap exists between domain scientists, computational scientists and AI-experts. A discussion of how the framework facilitates cross-disciplinary collaboration and integration across diverse technologies lower the risk in biosecurity would strengthen the work. 
3. Line 521, the authors mention “Current incentives for biotechnology firms and synthetic biology providers are misaligned as voluntary compliance with screening protocols often adds cost without tangible benefit” and lines #238-240 mentions "provide funding and cloud infrastructure/edge-computing tools to enable AI diagnostic use in low-resource settings," highlighting that capacity building for low- and middle-income countries is important for equitable implementation. Could the authors discuss or suggest what would be the best governance structure for resource allocation and decision making? 
4. The authors proposed global governance framework mentions "international incentives", the authors should also discuss how can this governance body enforce compliance across non-signatory nations and also what happens a top performer/country in AI technology refuse to adopt the global guidelines. 
5. In line 273, the authors acknowledge that "even though creation of viruses through synthetic biology is a WHO biorisk scenario, there is no unified or enforceable global mechanism," the authors could comment or discuss whether there have been previous attempts in this area and, if so, to provide an analysis of why those efforts were unsuccessful. 
6. It would be helpful if the authors could elaborate on the steps undertaken by the WHO within the Global Guidance Framework following the cited reference #47. 

Author Response

General Comment: "Overall, the review is well organized, easy to follow, and well motivated. The manuscript can be further improved if the authors could discuss and address the following suggestions and comments."

Response: We thank the reviewer for the positive assessment and have addressed all six specific comments below.

Comment 2.1: "Authors mention: 'A panel of experts should be formed to manually validate sequences flagged as synthetically altered, and output formats should be graphical and easily understood.' -- manually validating the sequences may also lead to human errors and higher expenses, inclusion of cost-benefit analysis of proposed framework would highly support the authors views."

Response: We appreciate this important point. We have added a cost-benefit analysis in Section 7 (lines 587-593). We acknowledge that manual expert validation incurs costs including financial overhead, time delays, and potential human error. However, we argue that the catastrophic risk of allowing highly pathogenic engineered sequences to bypass detection far exceeds these expenses. We propose a tiered urgency-based triage system that optimizes resources by reserving intensive scrutiny for high-risk cases while enabling verified research to proceed. We note that such systems could integrate into existing infrastructure like the WHO BioHub Initiative to distribute costs internationally (reference #103).

Comment 2.2: "Not all AI-generated sequences, structures, or compounds can be readily synthesized and experimentally validated. Additionally, a significant knowledge gap exists between domain scientists, computational scientists and AI-experts. A discussion of how the framework facilitates cross-disciplinary collaboration and integration across diverse technologies lower the risk in biosecurity would strengthen the work."

Response: Excellent point. We have addressed this in Section 7 (lines 598-609), where we discuss the need to bridge the knowledge divide between machine learning engineers and domain virologists. We propose integrated review structures where AI predictions are grounded in biological synthesizability and experimental validation. Specifically, we recommend standardized documentation of model limitations, interdisciplinary training programs pairing virologists with AI developers, and open platforms linking predictions with synthesizability databases. These mechanisms enhance both scientific quality and biosecurity by ensuring AI developers understand biological constraints while biologists understand AI limitations.

Comment 2.3: "Line 521, the authors mention 'Current incentives for biotechnology firms and synthetic biology providers are misaligned as voluntary compliance with screening protocols often adds cost without tangible benefit' and lines #238-240 mentions 'provide funding and cloud infrastructure/edge-computing tools to enable AI diagnostic use in low-resource settings,' highlighting that capacity building for low- and middle-income countries is important for equitable implementation. Could the authors discuss or suggest what would be the best governance structure for resource allocation and decision making?"

Response: We appreciate this request for specificity. While we discuss resource allocation mechanisms throughout Section 6 (particularly in subsection 5 on International Support Structures), we have strengthened the discussion in Section 7 (lines 614-615) by addressing the need for economic incentives. We propose "Certified Responsible Provider" designations, liability protections, and fast-track regulatory approvals that make biosafety compliance a competitive market asset. These mechanisms, combined with supply-chain controls on DNA synthesis equipment and cloud infrastructure, create both incentives and enforcement pathways. The specific governance structure would involve UN-WHO coordination modeled on successful frameworks like CEPI, with prioritization for regions with high epidemic risk and greatest infrastructure gaps.

Comment 2.4: "The authors proposed global governance framework mentions 'international incentives', the authors should also discuss how can this governance body enforce compliance across non-signatory nations and also what happens a top performer/country in AI technology refuse to adopt the global guidelines."

Response: This is a critical question we have now addressed in Section 7 (lines 610-622). We discuss enforcement mechanisms including supply-chain controls on DNA synthesis equipment and cloud infrastructure, combined with economic incentives that make biosafety compliance advantageous. For non-signatory nations, we propose leveraging the fact that major DNA synthesis providers and cloud computing resources are predominantly located in potential signatory nations, creating natural enforcement points. If major AI powers refuse adoption, we argue for a modular framework design that allows partial implementation—nations may coordinate on diagnostics and surveillance even while disagreeing on synthesis screening stringency. This flexibility increases the likelihood of at least partial buy-in from major players.

Comment 2.5: "In line 273, the authors acknowledge that 'even though creation of viruses through synthetic biology is a WHO biorisk scenario, there is no unified or enforceable global mechanism,' the authors could comment or discuss whether there have been previous attempts in this area and, if so, to provide an analysis of why those efforts were unsuccessful."

Response: Excellent suggestion for historical context. We have added analysis of previous governance attempts in Section 7 (lines 610-613). We discuss the Biological Weapons Convention (BWC), which lacks verification mechanisms following the 2001 protocol collapse over proprietary and security concerns (reference #106). We explain how our framework learns from these historical failures by incorporating supply-chain controls, economic incentives, and technical verification mechanisms rather than relying solely on voluntary compliance. This comparative analysis strengthens the rationale for our proposed approach.

Comment 2.6: "It would be helpful if the authors could elaborate on the steps undertaken by the WHO within the Global Guidance Framework following the cited reference #47."

Response: We have elaborated on WHO's implementation efforts in Section 7 (lines 616-618). We note that the WHO's 2022 Global Guidance Framework has been piloted in regions including Uganda during 2023-2024 (reference #107). These pilots provide valuable real-world testing of governance approaches. We use this as context to argue for the need for our more comprehensive framework with concrete implementation mechanisms and dedicated resources, building on WHO's foundational work.

Reviewer 3 Report

Comments and Suggestions for Authors

Well written paper. Some suggested improvements:

• Adding visuals and tables related to the research increases the visibility of the paper.
• While the research process has been defined well but no indication of future work
• Research limitation can have a separate section instead of just adding to the conclusion. 
• Can have a section explaining industrial benefit for adapting such research outcome.
• What are the existing industrial AI work trials related to virology?

 

Author Response

General Comment: "Well written paper. Some suggested improvements:"

Response: We thank the reviewer for the positive assessment and have addressed all suggestions below.

Comment 3.1: "Adding visuals and tables related to the research increases the visibility of the paper."

Response: Agreed. We have added three visual elements: Table 1 (lines 138-139) presenting a SWOT analysis of AI integration in global virology; Figure 1 (line 52) visually grouping AI tools by application area; and Figure 2 (line 546) showing the proposed governance framework as a flowchart. All are referenced in the main text with descriptive captions.

Comment 3.2: "While the research process has been defined well but no indication of future work"

Response: We have added Section 7 "Future Directions and Policy Limitations" which explicitly addresses future research directions (lines 598-622). We identify several critical areas including: evaluating WHO pilot programs' effectiveness, developing automated risk assessment tools for AI-generated sequences, conducting cost-benefit analyses of various governance structures, and exploring incentive mechanisms that align commercial interests with biosecurity objectives.

Comment 3.3: "Research limitation can have a separate section instead of just adding to the conclusion."

Response: Excellent suggestion. We have created Section 7 which begins with a dedicated limitations discussion (lines 581-597). We acknowledge that this framework remains conceptual pending empirical validation, that our analysis relies on literature current to 2025 while AI evolves rapidly, that manual validation introduces economic constraints, and that political will and international coordination challenges remain substantial.

Comment 3.4: "Can have a section explaining industrial benefit for adapting such research outcome."

Response: We have addressed industrial benefits in Section 7 (lines 602-609 and lines 614-615). We discuss how standardized frameworks reduce regulatory uncertainty and create competitive advantages through mechanisms like "Certified Responsible Provider" designations, liability protections, and fast-track regulatory approvals. We also note how integrated review structures and open platforms benefit industry by improving the quality and reliability of AI-driven research.

Comment 3.5: "What are the existing industrial AI work trials related to virology?"

Response: We have added specific industrial examples in Section 7 (lines 602-606). We cite Moderna's use of computational biology to design the mRNA-1273 vaccine in less than a day after the SARS-CoV-2 genome was released, with the first clinical trial beginning 66 days later (reference #104). We also discuss BenevolentAI's use of knowledge-graph analysis to identify baricitinib as a COVID-19 treatment within weeks of the first known cases (reference #105). These examples demonstrate AI's transformative potential in industrial virology applications.

Reviewer 4 Report

Comments and Suggestions for Authors

Overall, this paper addresses a critically important and fast-evolving topic at the intersection of AI, virology, and global security. The author demonstrates a sophisticated understanding of the technical landscape and the accompanying governance gap. The paper’s primary value lies in its synthesis of diverse evidence into coherent, actionable policy recommendations, making a compelling case for proactive and coordinated international action.

However, to achieve its full potential and meet the standards for publication, a major revision is required. The analysis, while thorough, remains at a high level and would benefit from deeper exploration of implementation pathways and counterarguments. The almost complete lack of visual aids (tables, figures) is a significant missed opportunity to clarify complex relationships and processes for the reader. Strengthening the paper's analytical depth, adding illustrative graphics, and extending the discussion of practical challenges will transform it from a valuable review into an essential and influential policy roadmap.

Detailed points to address within this major revision are:

-The paper provides a timely, comprehensive, and well-structured analysis of the dual-use potential of AI in virology, effectively balancing its benefits for public health with its significant biosecurity risks.

-The methodology is primarily a critical content analysis and policy review, drawing effectively on case studies, public databases, and documented applications to substantiate its arguments.

-The governance recommendations are a key strength, offering targeted, multi-stakeholder proposals (for governments, the UN, research institutions, and firms) that are both specific and adaptable for evolving technologies.

-The paper successfully connects historical precedents (gain-of-function research, synthetic virology) with contemporary AI capabilities, providing crucial context for the urgency of its proposed governance framework.

-The analysis would be significantly strengthened by the inclusion of original graphics (e.g., a conceptual model of the AI-virology dual-use dilemma, a flowchart of the proposed governance framework) and more detailed, illustrative tables to summarise tools, risks, and policy levers.

-The discussion, while broad, needs extension into a more nuanced analysis of the practical trade-offs, implementation challenges, and potential unintended consequences of the proposed regulations, particularly for scientific progress and global equity.

-Testing of the proposed ideas is inherently conceptual; however, the paper could bolster its case by including a brief comparative analysis of similar governance frameworks in other dual-use fields (cybersecurity, nuclear technology) or a SWOT analysis of its own recommendations.

-The writing is clear and authoritative, but some sections are quite dense. Improving the narrative flow and breaking down complex paragraphs would enhance readability for a multidisciplinary audience.

-The references are extensive and relevant, showcasing a strong command of the literature across virology, AI, synthetic biology, and international policy.

 

Author Response

General Comment: "Overall, this paper addresses a critically important and fast-evolving topic at the intersection of AI, virology, and global security. The author demonstrates a sophisticated understanding of the technical landscape and the accompanying governance gap. [...] However, to achieve its full potential and meet the standards for publication, a major revision is required."

Response: We are grateful for this thorough and constructive assessment. We have undertaken a major revision addressing all specific points raised below.

Comment 4.1: "The analysis would be significantly strengthened by the inclusion of original graphics (e.g., a conceptual model of the AI-virology dual-use dilemma, a flowchart of the proposed governance framework) and more detailed, illustrative tables to summarise tools, risks, and policy levers."

Response: We have added three visual elements as requested: Table 1 (lines 138-139) provides a SWOT analysis summarizing strengths, weaknesses, opportunities, and threats of AI integration in virology; Figure 1 (line 52) presents a conceptual model of AI's roles across diagnostic, surveillance, drug discovery, and synthetic biology domains; and Figure 2 (line 546) provides a detailed flowchart of the proposed governance framework showing the relationships between international bodies, advisory panels, policymakers, and end users. All visuals are referenced in the main text with comprehensive captions.

Comment 4.2: "The discussion, while broad, needs extension into a more nuanced analysis of the practical trade-offs, implementation challenges, and potential unintended consequences of the proposed regulations, particularly for scientific progress and global equity."

Response: We have added Section 7 which provides exactly this nuanced analysis (lines 581-622). We address practical constraints including the economic burden of manual validation, political will challenges, and international coordination difficulties. We discuss trade-offs such as the cost of validation versus catastrophic biosecurity risks, and propose tiered triage systems to balance these competing concerns. We also address equity by noting the importance of distributing costs through international infrastructure like the WHO BioHub Initiative and by proposing economic incentives that align industry interests with public health goals.

Comment 4.3: "Testing of the proposed ideas is inherently conceptual; however, the paper could bolster its case by including a brief comparative analysis of similar governance frameworks in other dual-use fields (cybersecurity, nuclear technology) or a SWOT analysis of its own recommendations."

Response: Excellent suggestions, both of which we have implemented. We have added Table 1 (lines 138-139) providing a comprehensive SWOT analysis of the proposed framework. Additionally, we have added comparative analysis in Section 7 (lines 610-613) discussing the Biological Weapons Convention and its historical failures due to lack of verification mechanisms. We explain how our framework learns from these precedents by incorporating technical verification, supply-chain controls, and economic incentives rather than relying solely on voluntary compliance.

Reviewer 5 Report

Comments and Suggestions for Authors

Dear Author(s), 

 

I do believe that your manuscript is highly relevant for the AI body of literature and for the Virology specialty. I do not have anything to add regarding the content of the manuscript but rather to the structure of the study. First, I thought it is going to be a narrative review or a review, but I did not find anything related to them. Is has the structure of a review but there are some sections missing. For instance, a minimum methodology should be included, in which the research questions are included, the used databases for data collection and the keywords you have used while structuring your critical content analysis. An explanation of how you employed the content analysis will be highly recommended. 

So, we have a very good content analysis review, what is the next step? A Discussion section should be included as well, by pointing out the main topics, as well as the limitations of the study and the further research directions. Who and how can use the information you included in the study?

 

Thank you and good luck!

Author Response

General Comment: "I do believe that your manuscript is highly relevant for the AI body of literature and for the Virology specialty. I do not have anything to add regarding the content of the manuscript but rather to the structure of the study."

Response: We appreciate this positive assessment of the content and have addressed all structural concerns below.

Comment 5.1: "First, I thought it is going to be a narrative review or a review, but I did not find anything related to them. [...] For instance, a minimum methodology should be included, in which the research questions are included, the used databases for data collection and the keywords you have used while structuring your critical content analysis. An explanation of how you employed the content analysis will be highly recommended."

Response: We have added a methodology section in the Introduction (lines 72-86) that addresses all these points. We clarify that this is a critical content analysis (not a systematic review) and specify: (1) the core objectives guiding our literature selection (identifying high-impact AI tools, assessing dual-use potential, evaluating oversight feasibility); (2) databases consulted (PubMed, bioRxiv, arXiv); (3) timeframe (primarily 2018-2025); (4) search focus (intersection of AI-assisted virology, biosecurity, and generative protein design); and (5) our thematic synthesis approach organizing findings into four functional domains (diagnostics, surveillance, drug discovery, synthetic design).

Comment 5.2: "So, we have a very good content analysis review, what is the next step? A Discussion section should be included as well, by pointing out the main topics, as well as the limitations of the study and the further research directions. Who and how can use the information you included in the study?"

Response: Excellent point. We have added Section 7 "Future Directions and Policy Limitations" which serves as the Discussion section requested. This section addresses: (1) study limitations including the conceptual nature of the framework, reliance on current literature, economic constraints of validation, and political coordination challenges (lines 581-597); (2) future research directions including evaluation of WHO pilots, development of automated risk assessment tools, and exploration of incentive mechanisms (lines 598-622); and (3) practical implementation pathways including cross-disciplinary collaboration structures and learning from historical governance failures. This section explicitly addresses who can use this framework (international bodies, national governments, industry, research institutions) and how (through the specific policy mechanisms detailed throughout the manuscript).

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have satisfactorily addressed both comments by adding clear and informative figures. The visual summaries improve readability and help convey both the scope of AI applications and the proposed governance framework. The revisions are acceptable, and no further changes are required.

Author Response

“The authors have satisfactorily addressed both comments by adding clear and informative figures. The visual summaries improve readability and help convey both the scope of AI applications and the proposed governance framework. The revisions are acceptable, and no further changes are required.”

We thank the reviewer for their approval and their helpful comments from the first submission.

Reviewer 4 Report

Comments and Suggestions for Authors

Some of my round 1 recommendations and remarks were addressed. However, the paper still lacks critical analysis and depth.

These points require further work:

Comment 4.2: "The discussion, while broad, needs extension into a more nuanced analysis of the practical trade-offs, implementation challenges, and potential unintended consequences of the proposed regulations, particularly for scientific progress and global equity."

Comment 4.3: "Testing of the proposed ideas is inherently conceptual; however, the paper could bolster its case by including a brief comparative analysis of similar governance frameworks in other dual-use fields (cybersecurity, nuclear technology) or a SWOT analysis of its own recommendations."

Author Response

“These points require further work:

Comment 4.2: "The discussion, while broad, needs extension into a more nuanced analysis of the practical trade-offs, implementation challenges, and potential unintended consequences of the proposed regulations, particularly for scientific progress and global equity."”

We appreciate the reviewers effort to enrich the manuscript and policy proposals, and have substantially expanded the Limitations section (lines 556–580) to provide the analysis requested. This expanded discussion now explicitly addresses:

Trade-offs for scientific progress: We acknowledge that mandatory validation and screening protocols may slow beneficial innovation, with drug discovery programs potentially missing critical pandemic response windows and diagnostic tools being delayed from reaching clinics (lines 558–561). We characterize these as "real trade-offs between security and scientific velocity," not merely hypothetical concerns.

Unintended consequences for global equity: We discuss how stringent requirements may disproportionately burden smaller firms and under-resourced institutions in low- and middle-income countries, potentially widening rather than narrowing global research capacity gaps (lines 562–566). We note that "a framework optimized in a wealthy location but unaffordable in a developing region becomes a mechanism for exclusion rather than protection" (lines 566–567), and that frameworks perceived as Western regulatory control risk provoking resistance from nations whose participation is most critical (lines 568–571).

Risk-proportionate governance: We explicitly acknowledge that overly stringent governance may drive research into less transparent settings or discourage beneficial applications that outweigh their misuse potential (lines 577–580). We emphasize the need for interventions that are "risk-proportionate, economically feasible, and politically achievable across vastly different national contexts" (lines 575–577).

This expanded analysis addresses with the genuine tensions inherent in biosecurity governance, treating them as substantive implementation challenges rather than peripheral caveats.

“Comment 4.3: "Testing of the proposed ideas is inherently conceptual; however, the paper could bolster its case by including a brief comparative analysis of similar governance frameworks in other dual-use fields (cybersecurity, nuclear technology) or a SWOT analysis of its own recommendations."”

We have expanded the discussion of trade-offs:

SWOT analysis: Table 1 (lines 134–138) provides a comprehensive SWOT analysis of AI integration in global virology, systematically evaluating Strengths (accelerated antiviral discovery, protein folding predictions, real-time surveillance, synthetic modification detection), Weaknesses (black-box decision-making, dataset dependency, computational costs, interpretability challenges), Opportunities (UN-led biosecurity clearinghouse, standardized screening, environmental metagenomics integration, bridging diagnostic gaps), and Threats (democratization of dual-use design, AI outpacing policy, generative model hallucinations, private sector misalignment).

Comparative analysis with other dual-use governance regimes: We have substantially expanded Section 7 (lines 642–669) to include explicit comparative analysis drawing on three governance precedents:

1. Nuclear governance (IAEA): Lines 642–655 analyze how IAEA safeguards achieved sustained international compliance through technical verification mechanisms (mandatory reporting, inspection protocols), benefit-sharing programs (civilian nuclear power access), and graduated sanctions. We explain how our framework incorporates analogous features: automated DNA synthesis screening logs function as technical verification, while economic incentives (certification benefits, fast-track approvals) create tangible participation rewards.
2. Cybersecurity governance (Budapest Convention): Lines 656–669 examine how the Budapest Convention's limited adoption beyond Europe/North America stems from sovereignty concerns and inadequate involvement of major cyber powers (China, Russia, India) in treaty drafting. We extract specific lessons: frameworks developed primarily by Western nations risk legitimacy deficits, and rigid all-or-nothing structures create adoption barriers. Our UN-led, modular approach directly responds to these failures by prioritizing inclusive multilateral development and allowing partial adoption without full treaty ratification.
3. Biological governance (BWC): Lines 629–632 discuss the Biological Weapons Convention's lack of verification mechanisms following the 2001 protocol collapse, demonstrating the inadequacy of voluntary compliance frameworks.

Each comparison extracts specific design lessons that informed our framework's architecture, demonstrating how AI biosecurity governance can learn from both successes (IAEA) and failures (BWC, Budapest Convention) in other dual-use domains.

Reviewer 5 Report

Comments and Suggestions for Authors

All my improvement suggestions have been addressed. 

 

Thank you!

Author Response

“All my improvement suggestions have been addressed. 

Thank you!”

We thank the reviewer for their helpful comments from the first round and are delighted that they approve of our manuscript.