Round 1

Reviewer 1 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

Dear Authors,

Thank you for submitting this engaging and well-structured manuscript. This study addresses a highly relevant topic, microbiome functional signatures across BMI categories in the MENA region, and the combined taxonomic, functional, and covariance analyses provide valuable insights. The manuscript is clear, the figures are informative, and the data analysis is generally well performed. Below are suggestions to further strengthen the scientific clarity and interpretability of the study.

The abstract highlights Lachnospira, Lactobacillus, and Roseburia as enriched in non-obese individuals. While these taxa do appear in Figure 2, the Results section does not describe or discuss these findings in detail. To ensure consistency, please expand the Results narrative to match the taxa emphasized in the abstract.

I suggest adding a brief clarification in the methods regarding: i) the exact BMI cutoff used to define the two groups, ii) whether BMI was treated as a categorical or continuous variable in the DESeq2 model, and iii) whether any normalization or rarefaction was applied before alpha and beta diversity analyses. These additions would help improve clarity.

I suggest addressing the age imbalance shown in Table 1, as participants in the higher BMI category appear to be older on average. Since age is already included as a covariate in the DESeq2 model, adding a brief explanation of whether age remained significant in the analysis would help readers understand its impact on the results. If applicable, you may also mention whether any additional sensitivity checks (such as age-matched subsampling or alternative adjustments) were considered. Including this minor clarification would certainly strengthen confidence in the robustness of the microbiome-BMI associations reported in the study.

I also recommend adding a brief note in the Discussion regarding the interpretation of the KEGG pathways enriched in the obese cohort, as several of these pathways are associated with infection- or pathogenic-related functions. Since PICRUSt2 predictions are based on 16S-derived inference rather than direct metagenomic measurements, it would be helpful to acknowledge the inherent limitations of this approach, including the potential for pathway overrepresentation due to biases in available reference genomes. A brief clarification would ensure readers interpret these functional predictions appropriately and further strengthen the overall discussion.

Thank you again for your work. I look forward to the revised version.

Author Response

We are grateful for the reviewer's insight. A more thorough explanation of the taxa Lachnospira, Lactobacillus, and Roseburia has been included in the Results section, emphasizing their relative richness in non-obese individuals as shown in Figure 2 and newly added Figures S9 and S10. This addition guarantees that the Abstract and the Results are in line.]

We appreciate the reviewer's recommendation. The BMI limit for group categorization, the use of BMI as a categorical variable in DESeq2, and the fact that no rarefaction was performed prior to diversity analysis because standard normalization procedures were employed within the Phyloseq framework are all now clearly stated in the Methods section.

We appreciate this thoughtful suggestion. Although age was a covariate in the DESeq2 model, it did not show up as a significant predictor of microbial abundance, as we have clarified in the Results and Discussion. Additionally, we have seen that age-matched subgroups used in sensitivity analyses yielded consistent results, supporting the robustness of our conclusions shown in Figures S4-S8.

We appreciate the reviewer's valuable input. To recognize the limitations of PICRUSt2-based functional inference and to direct the understanding of KEGG pathway data appropriately, we have included a clarification in the Discussion, “The KEGG pathways found to be enriched in obese individuals many of which are associated with pathogenic and infection-related mechanisms are based on 16S rRNA-derived inference through PICRUSt2 rather than direct metagenomic sequencing, so it is crucial to interpret the functional predictions cautiously. Therefore, insufficient functional annotation or biases in the reference genome may affect these predictions. Future research using metabolomics or shotgun metagenomics will be useful for confirming these deduced functional relationships”.

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

Thank you for the opportunity to review “Beyond Diversity: Functional Microbiome Signatures Linked to Obesity.” The analysis of gut microbiome variations across BMI categories is promising. The comprehensive use of network analysis and PICRUSt2 functional profiling is particularly strong, and the focus on the MENA region addresses an important gap in the literature. However, I have a few constructive suggestions that I believe will elevate this already promising manuscript:

Major Comments:

While the introduction effectively outlines the gut microbiome’s role in metabolism, it would benefit from a deeper discussion of the unique epidemiological context of obesity in the MENA region. Currently, it does not fully connect global microbiome research with the region-specific metabolic health challenges that make this study’s approach distinctive and potentially impactful.
While Table 1 provides a basic overview, the manuscript would benefit from more comprehensive participant profiling. It would be helpful to expand the demographic data to include dietary patterns, physical activity levels, and additional metabolic health indicators beyond BMI.
The study’s methodological limitations primarily arise from the small and unbalanced sample size (103 total: 75 non-obese vs. 28 obese). This reduces statistical power, may introduce bias, and warrants cautious interpretation of the microbiome findings, along with a clear justification of the sample size selection.
The DNA extraction and sequencing methods are well-described. I recommend including more details about the quality control steps to ensure the reliability and reproducibility of the sequencing data.
The discussion of Lachnospira and Phascolarctobacterium is intriguing. It would be helpful if the authors could provide more speculation on the potential mechanistic links between these bacterial genera and metabolic processes.
The statistical approaches are robust; however, a brief discussion of the limitations of 16S rRNA gene-based predictions would further strengthen the manuscript’s scientific rigor.
The covariance network analysis is sophisticated. I recommend adding a schematic or simplified diagram to help readers who may be less familiar with such complex visualizations.
It would be valuable to discuss how the study findings could translate into potential therapeutic or dietary interventions

Author Response

While the introduction effectively outlines the gut microbiome’s role in metabolism, it would benefit from a deeper discussion of the unique epidemiological context of obesity in the MENA region. Currently, it does not fully connect global microbiome research with the region-specific metabolic health challenges that make this study’s approach distinctive and potentially impactful.

We appreciate the reviewer's recommendation. In order to highlight how this regional focus enhances the study's contribution, we have added more epidemiological context regarding obesity rates in the MENA region to the Introduction.

While Table 1 provides a basic overview, the manuscript would benefit from more comprehensive participant profiling. It would be helpful to expand the demographic data to include dietary patterns, physical activity levels, and additional metabolic health indicators beyond BMI.

We acknowledge this issue and have added information on self-reported physical activity, dietary fiber intake habits, and certain metabolic health markers to the demographic profiling, as shown in Figure S4-8.

The study’s methodological limitations primarily arise from the small and unbalanced sample size (103 total: 75 non-obese vs. 28 obese). This reduces statistical power, may introduce bias, and warrants cautious interpretation of the microbiome findings, along with a clear justification of the sample size selection.

We agree with the reviewer. The sample size has been justified in light of data availability and regional recruiting limits, and a note acknowledging this limitation has been added.

The DNA extraction and sequencing methods are well-described. I recommend including more details about the quality control steps to ensure the reliability and reproducibility of the sequencing data.

Details on the quality control procedures used during sequence processing have been added to the methods as shown in Figure S1-3.

The discussion of Lachnospira and Phascolarctobacterium is intriguing. It would be helpful if the authors could provide more speculation on the potential mechanistic links between these bacterial genera and metabolic processes.

Potential metabolic pathways that connect these species to host physiology have been explained in detail in the Discussion, “Mechanistically, Lachnospira promotes energy balance by producing butyrate and propionate, which control the integrity of the intestinal barrier and appetite signals through the GLP-1 and PYY pathways. Hepatic gluconeogenesis and lipid metabolism, on the other hand, may be impacted by Phascolarctobacterium through the production of propionate and the use of succinate. Their conflicting links with obesity may be partially explained by these complimentary yet different metabolic roles.”

The statistical approaches are robust; however, a brief discussion of the limitations of 16S rRNA gene-based predictions would further strengthen the manuscript’s scientific rigor.

The limitations of 16S rRNA functional inference have been clarified in the discussion as a limitation.

The covariance network analysis is sophisticated. I recommend adding a schematic or simplified diagram to help readers who may be less familiar with such complex visualizations.

We are grateful for this recommendation. To aid in interpretation, a simplified schematic has been included as an additional figure S11.

It would be valuable to discuss how the study findings could translate into potential therapeutic or dietary interventions

We appreciate the reviewer and have added to the discussion.

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

I read with interest the manuscript titled "Beyond Diversity: Functional Microbiome Signatures Linked to Obesity."

The phrase "potential protective effects" is somewhat vague.

The evidence supporting the protective effects of these taxa is often associative, not necessarily causal.

The term "Lactobacilli" is plural but used as a genus name (Lactobacillus).

The causal role of Lactobacillus in reducing adiposity and insulin resistance is often based on associative studies; causality is not firmly established. The evidence is sometimes strain-specific; generalizing across the genus can be misleading.

While Lachnospira can produce SCFAs, the extent of its contribution relative to other taxa is variable and context-dependent. The statement "significantly enhances gut health" might be overstated; it’s more accurate to say it may contribute.

"Tightly interconnected microbial network" is speculative unless supported by specific network analyses.

The sample size (n=126) may be small for microbiome studies, which often require larger cohorts to detect meaningful differences, especially considering variability in microbiota composition.

No mention of how participants were recruited or inclusion/exclusion criteria, which affects reproducibility and bias assessment.

Lack of details on sample stabilization (e.g., use of preservatives, time between collection and freezing) which impacts microbiome integrity.

No mention of whether samples were collected at home or in clinical settings, and whether participants followed standardized protocols.

Include DNA concentration thresholds for inclusion in sequencing.

Consider reporting DNA integrity metrics (e.g., fragment size distribution).

The V3-V4 regions are common, but the primer pair used can influence taxonomic resolution; specify the exact primer sequences.

No mention of sequencing depth (reads per sample), which affects diversity estimates and statistical power.

No mention of negative controls or mock communities to assess contamination or sequencing bias.

No mention of quality filtering thresholds, chimera removal specifics, or truncation lengths.

No description of how ambiguous or low-confidence taxonomic assignments were handled.

Provide specific DADA2 parameters (e.g., truncLen, maxEE, truncQ).

Clarify the version of DADA2 used.

Describe steps taken to ensure data quality.

SILVA database versions can vary; specify the exact version used (e.g., SILVA v138).

Taxonomic resolution at genus level may be limited; mention if species-level assignments were attempted or possible.

No mention of rarefaction depth or whether data were normalized prior to diversity calculation.

For beta diversity, mention if samples were rarefied or normalized.

DESeq2 is designed for count data but assumes certain distributions; microbiome data are compositional and may require alternative methods (e.g., ANCOM, ALDEx2).

Using DESeq2 without appropriate compositional data transformations can lead to misleading results. Justify the choice of DESeq2 or consider complementary methods suitable for compositional data.

Clarify how data were normalized or transformed prior to DESeq2 analysis.

Clarify the normalization process used before pathway prediction.

Discuss limitations of functional inference from 16S data.

The choice of statistical model for pathway data should be justified, especially given the compositional nature of inferred functional data.

No mention of multiple testing correction (e.g., FDR correction).

Address how zeros were managed prior to CLR (e.g., imputation).

Consider alternative correlation methods suited for compositional data (e.g., SparCC).

Clarify whether the network analysis was exploratory or hypothesis-driven.

After reviewing the Introduction and Materials and Methods sections, I believe that a thorough revision of these parts is essential before proceeding to evaluate the remaining sections. Ensuring clarity, methodological rigor, and comprehensive detail at this stage will strengthen the overall quality of your study and enhance its reproducibility and scientific validity.

A careful revision focusing on precise descriptions of the study population, sample collection procedures, sequencing protocols, bioinformatics pipelines, and statistical analyses will provide a solid foundation for the interpretation of your results. Once these sections are thoroughly refined, a more meaningful and constructive evaluation of the subsequent parts of your manuscript can be confidently undertaken.

Author Response

The phrase "potential protective effects" is somewhat vague.

The evidence supporting the protective effects of these taxa is often associative, not necessarily causal.

The term "Lactobacilli" is plural but used as a genus name (Lactobacillus).

We agree and have updated the introduction to make it clear that the majority of the evidence connecting these taxa to metabolic advantages is associative. We have also updated the language for Lactobacillus.

"Tightly interconnected microbial network" is speculative unless supported by specific network analyses.

We have changed the comments to use more careful wording and appreciate the reviewer's clarification.

The sample size (n=126) may be small for microbiome studies, which often require larger cohorts to detect meaningful differences, especially considering variability in microbiota composition.

No mention of how participants were recruited or inclusion/exclusion criteria, which affects reproducibility and bias assessment.

Acknowledging this, we have included information about participant recruitment, inclusion/exclusion criteria, and the rationale for the sample size in the Methods section.

Lack of details on sample stabilization (e.g., use of preservatives, time between collection and freezing) which impacts microbiome integrity.

No mention of whether samples were collected at home or in clinical settings, and whether participants followed standardized protocols.

Include DNA concentration thresholds for inclusion in sequencing.

Consider reporting DNA integrity metrics (e.g., fragment size distribution).

The V3-V4 regions are common, but the primer pair used can influence taxonomic resolution; specify the exact primer sequences.

We have added more information regarding DNA quality control and sample handling as shown in Methods and Figures S1-3.

No mention of sequencing depth (reads per sample), which affects diversity estimates and statistical power.

No mention of negative controls or mock communities to assess contamination or sequencing bias.

No mention of quality filtering thresholds, chimera removal specifics, or truncation lengths.

No description of how ambiguous or low-confidence taxonomic assignments were handled.

For reproducibility and transparency, we have supplied these details in Methods at Sequencing of the Bacterial 16S rRNA Genes.