Integrative Analysis of Volatile Flavor Compounds and Transcriptome Reveals Underlying Mechanisms Linked to Fatty Acid Content in Dabieshan Cattle
, Qian Li 1,†, Hai Jin 1, Shuanping Zhao 1, Huibin Zhang 1, Xinyi Du 1, Qinggang Li 1 and Lei Xu 1,*Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis manuscript presents an integrative multi-omics analysis combining fatty acid profiling, volatile compound analysis (GC×GC-TOF-MS), and transcriptomics to investigate flavor formation mechanisms in Dabieshan cattle. The topic is relevant to meat science and molecular breeding, and the integration of lipidomics and transcriptomics is a clear strength of the study. However, several methodological and interpretative issues need to be addressed before the manuscript can be considered for publication.
Major Comments
1. Sample Size and Statistical Power
The study is based on 20 animals, but only 8 individuals (4 per group) were selected for transcriptomic and multi-omics integration analyses. This extremely small sample size substantially limits statistical power and increases the risk of both false positives and false negatives.
Please justify the sample size statistically.
Provide power analysis or effect size estimation.
Discuss the limitation more explicitly in the Discussion section.
For high-impact omics studies, n = 4 per group is generally considered underpowered.
2. Grouping Strategy Based on Total Fatty Acid Content
The classification into H and L groups is based solely on total fatty acid content. However:
Total fatty acid content is not necessarily equivalent to differences in specific fatty acid profiles.
Please clarify whether grouping was performed before or after fatty acid profiling.
Provide clear statistical thresholds used for selecting “extreme” individuals.
Explain whether other confounding variables (carcass weight, IMF, diet variability, batch effects) were controlled.
This grouping strategy needs stronger justification.
3. Transcriptomic Analysis Concerns
Several methodological details are insufficient:
Was batch correction applied?
Were biological replicates independently sequenced?
How was multiple testing correction performed (FDR method explicitly stated)?
Why was edgeR selected instead of DESeq2? Provide rationale.
The PCA variance explanation (PC1 = 11.8%, PC2 = 29.1%) seems unusual (PC2 > PC1). Please clarify.
Additionally, enrichment analysis via DAVID is somewhat outdated. Consider validating with more recent tools (e.g., clusterProfiler).
4. Correlation Analysis and Causality Claims
The manuscript suggests regulatory mechanisms linking SGPL1, KLF15, and SLC27A6 to fatty acids and VOCs. However:
Only Spearman correlations were performed.
Correlation does not demonstrate causation.
No network inference, mediation analysis, or experimental validation was conducted.
Statements such as “constructed regulatory relationships” should be toned down. Please revise to indicate associative relationships rather than confirmed regulatory mechanisms.
5. ROAV Interpretation
(E)-2-Nonenal is identified as the key compound (ROAV = 100). However:
Odor threshold values used for ROAV calculation are not clearly cited.
Were thresholds species-specific or taken from literature?
Provide formula and reference source for ROAV calculation.
This is essential for reproducibility.
6. Biological Interpretation of SLC27A6
The interpretation of SLC27A6 is contradictory:
It is negatively correlated with fatty acids.
Yet positively correlated with VOCs derived from those fatty acids.
This paradox requires clearer mechanistic explanation or a more cautious interpretation.
Minor Comments
Numerous grammatical errors and awkward phrasing are present (e.g., verb tense inconsistencies, article usage, punctuation).
“KEEG” should be corrected to “KEGG”.
Some references appear inconsistently formatted.
Figure legends require clearer statistical annotations (exact p-values instead of only * or **).
OPLS-DA validation metrics (R², Q², permutation test) should be provided.
Clarify whether raw RNA-seq data are deposited in a public repository (GEO/SRA). Current data availability statement limits reproducibility.
Strengths of the Study
Integration of lipidomics, volatilomics, and transcriptomics.
Use of GC×GC-TOF-MS for comprehensive VOC profiling.
Focus on indigenous cattle breed with breeding relevance.
Identification of candidate genes linked to fatty acid metabolism.
Overall Recommendation: Major Revision
The study has scientific merit and novelty, but substantial methodological clarification, improved statistical rigor, and more cautious interpretation of regulatory mechanisms are required before publication.
Comments on the Quality of English LanguageThe manuscript is generally understandable; however, the English language requires substantial improvement to ensure clarity, precision, and scientific rigor. Numerous grammatical errors, inconsistent verb tenses, article misuse, awkward phrasing, and punctuation issues are present throughout the text. In several sections (particularly the Results and Discussion), long and complex sentences reduce readability and occasionally obscure the intended meaning.
In addition, some technical descriptions (e.g., transcriptomic analysis, statistical procedures, and interpretation of correlations) would benefit from clearer and more concise wording. Minor typographical errors (e.g., inconsistent abbreviations, spelling issues such as “KEEG” instead of “KEGG”) should also be corrected.
Professional English language editing by a native speaker or a specialized scientific editing service is strongly recommended prior to publication.
Author Response
Comments 1: 1. Sample Size and Statistical Power
The study is based on 20 animals, but only 8 individuals (4 per group) were selected for transcriptomic and multi-omics integration analyses. This extremely small sample size substantially limits statistical power and increases the risk of both false positives and false negatives.
Please justify the sample size statistically.
Provide power analysis or effect size estimation.
Discuss the limitation more explicitly in the Discussion section.
For high-impact omics studies, n = 4 per group is generally considered underpowered.
Response 1: Thank you for raising the important issue regarding sample size. We fully agree that n = 4 per group is generally considered relatively small. The Dabieshan cattle used in this study are a local livestock genetic resource in Anhui Province with a limited population size, and strict genetic resource protection policies combined with practical farming conditions substantially restricted the number of samples that could be collected through slaughter, ultimately yielding 20 individuals with consistent genetic backgrounds.To maximize detection efficiency under this limited sample size, we employed an extreme phenotype selection strategy, in which fatty acid content was first measured in all 20 samples, and the four individuals with the highest content (H group) and the four with the lowest content (L group) were selected for further analysis.This approach maximizes between-group differences and increases effect size, thereby improving statistical power with a smaller sample size. To further address the reviewer’s concern regarding statistical power,we have also explicitly discussed the limitations associated with small sample size in the Discussion section.
Comments 2: Grouping Strategy Based on Total Fatty Acid Content
The classification into H and L groups is based solely on total fatty acid content. However:
Total fatty acid content is not necessarily equivalent to differences in specific fatty acid profiles.
Please clarify whether grouping was performed before or after fatty acid profiling.
Provide clear statistical thresholds used for selecting “extreme” individuals.
Explain whether other confounding variables (carcass weight, IMF, diet variability, batch effects) were controlled.
This grouping strategy needs stronger justification.
Response 2: Thank the reviewer for the valuable comments on the grouping strategy. We fully agree that total fatty acid content is not equivalent to differences in specific fatty acid profiles. In this study, total fatty acid content was first calculated as the sum of all individual fatty acids measured by GC. Based on this calculated total fatty acid content, the extreme grouping (upper and lower 25%) was used to select individuals with overall differences in fatty acid levels. Subsequently, we compared the contents of 46 individual fatty acids between the two groups. The fatty acid profiles of the H and L groups are presented in Supplementary Table S1.We will add this clarification to the revised manuscript to better justify the grouping strategy.
Key confounding variables were strictly controlled during sample collection: all samples were obtained from Dabieshan cattle with consistent genetic backgrounds, age, approximately 350 kg body weight, and identical feeding conditions, effectively minimizing the potential influence of genetic background, age, carcass weight, and dietary variability on fatty acid metabolism. In addition, all samples were processed in a single batch to avoid batch effects. The body weight of Dabieshan cattle has been added in the revised manuscript.
Comments 3: Transcriptomic Analysis Concerns
Several methodological details are insufficient:
Was batch correction applied?
Were biological replicates independently sequenced?
How was multiple testing correction performed (FDR method explicitly stated)?
Why was edgeR selected instead of DESeq2? Provide rationale.
The PCA variance explanation (PC1 = 11.8%, PC2 = 29.1%) seems unusual (PC2 > PC1). Please clarify.
Additionally, enrichment analysis via DAVID is somewhat outdated. Consider validating with more recent tools (e.g., clusterProfiler ).
Response 3: We thank the reviewer for pointing out the errors in our manuscript. A total of eight individuals, comprising the high fatty acid group (Group H, n = 4) and low fatty acid group (Group L, n = 4), were selected for transcriptome sequencing. Each biological replicate was independently constructed as a separate library (no pooling), and all libraries were prepared in a single batch and subjected to paired-end sequencing (PE150) on the Illumina NovaSeq 6000 platform.
Differential expression analysis was performed using the edgeR software package. Considering the relatively small sample size in this study (n = 4 per group), edgeR was chosen over DESeq2 because it provides more robust dispersion estimation under small-sample conditions and more effectively controls the false positive rate. The description of the transcriptome methodology in the original manuscript was incorrect; the threshold has now been corrected to p < 0.05. Differentially expressed genes were identified based on the following criteria: |logâ‚‚FC| > 1 and p < 0.05, where logâ‚‚FC represents the log2-transformed fold change in gene expression and p represents the significance level from the differential expression test. These thresholds were established based on commonly used criteria for screening differentially expressed genes in transcriptomic studies, allowing for effective control of the false positive rate while identifying biologically meaningful differentially expressed genes.
We thank the reviewer for this insightful observation. Upon re-examining the multivariate analysis in our original manuscript, we found an inadvertent error: the plot described as PCA was actually an OPLS-DA score plot. When we further validated the OPLS-DA model, the Q² value was found to be negative, indicating that the model was invalid. Therefore, we have removed OPLS-DA results from the transcriptomic analysis in the revised manuscript. To provide an alternative visualization, we performed a genuine principal component analysis (PCA) on the same dataset. However, the separation between the H and L groups in the PCA plot was not pronounced, likely due to the small sample size (n=4 per group). As a result, we have placed this PCA plot in the supplementary materials (Supplementary Figure 1) without making any strong claims based on it. In the main text, we now rely on the statistically significant differences in total fatty acid content (P < 0.05) and the clear hierarchical clustering (Fig. 1C) to justify the grouping.We sincerely apologize for the confusion caused by this error and greatly appreciate the reviewer's careful reading and constructive suggestions, which have helped us improve the accuracy and transparency of our manuscript.
Thank you for your valuable suggestion. We understand the reviewer‘s concern regarding the use of updated analytical tools. Although DAVID is a relatively early-stage functional enrichment analysis tool, its development team has continuously maintained and updated it. Notably, at the end of 2021, DAVID underwent a major reconstruction and upgrade, significantly expanding its knowledgebase and species coverage, and it remains one of the widely used and reliable platforms in the field of functional enrichment analysis. We also noted that several recent transcriptomic studies have employed DAVID for functional enrichment analysis and have been published in high-quality journals. For example, Cheng et al. (2025) used DAVID for GO and KEGG enrichment analyses in their integrated network toxicology and transcriptomics study published in Ecotoxicology and Environmental Safety, and Romero López et al. (2025) also performed GO and KEGG enrichment analyses on the DAVID platform in their transcriptomic study of cervical cancer cells published in Computers in Biology and Medicine. We greatly appreciate the reviewer’s suggestion and fully recognize the advantages of using more up to date tools such as clusterProfiler for functional annotation and visualization. In future studies, we will actively consider adopting these tools to further enhance the novelty and completeness of our analyses.
Comments 4: Correlation Analysis and Causality Claims
The manuscript suggests regulatory mechanisms linking SGPL1, KLF15, and SLC27A6 to fatty acids and VOCs. However:
Only Spearman correlations were performed.
Correlation does not demonstrate causation.
No network inference, mediation analysis, or experimental validation was conducted.
Statements such as “constructed regulatory relationships” should be toned down. Please revise to indicate associative relationships rather than confirmed regulatory mechanisms.
Response 4: Thank you for your valuable comment. We fully agree that correlation analysis does not imply causation, and the wording regarding “regulatory mechanisms” in the original manuscript was indeed imprecise. In response to your suggestion, we have toned down causal statements such as “constructed regulatory relationships” throughout the manuscript, and all corresponding expressions have been revised to reflect associative relationships.
Comments 5: ROAV Interpretation
(E)-2-Nonenal is identified as the key compound (ROAV = 100). However:
Odor threshold values used for ROAV calculation are not clearly cited.
Were thresholds species-specific or taken from literature?
Provide formula and reference source for ROAV calculation.
This is essential for reproducibility.
Response 5:Thank you for your valuable comments, which are of great importance in improving the rigor and reproducibility of our manuscript. We have added references for the odor thresholds and the ROAV calculation formula in the Materials and Methods section. In addition, the odor threshold for (E)-2-nonenal was sourced from van Gemert L.J., Odor Thresholds: Compilations of Odor Threshold Values in Air, Water and Other Media, to ensure data traceability.
Comments 6: Biological Interpretation of SLC27A6
The interpretation of SLC27A6 is contradictory:
It is negatively correlated with fatty acids.
Yet positively correlated with VOCs derived from those fatty acids.
This paradox requires clearer mechanistic explanation or a more cautious interpretation.
Response 6:We thank the reviewer for pointing out this seemingly contradictory phenomenon, and we fully understand the concern. Regarding the issue that SLC27A6 expression is negatively correlated with fatty acid content but positively correlated with fatty acid-derived flavor compounds, we have attempted to provide an explanation from the perspective of the biological function of SLC27A6 in the manuscript.
SLC27A6 belongs to the fatty acid transport protein family, and its primary function is to mediate the transmembrane uptake and intracellular transport of long-chain fatty acids, rather than promoting fatty acid storage. Previous studies have shown that after facilitating the entry of fatty acids into cells, members of this protein family often direct them toward metabolic pathways such as β-oxidation. In addition, the expression level of fatty acid transport proteins is closely associated with the oxidation rate of palmitate. Based on this, we speculate that high expression of SLC27A6 may accelerate the uptake and oxidative metabolism of fatty acids, thereby reducing their net content while promoting the generation of their oxidative products. In the present study, the higher expression of SLC27A6 in the low fatty acid group (L group) may lead to a faster entry of the aforementioned precursor fatty acids into oxidative pathways, resulting in decreased fatty acid content; meanwhile, since fatty acid oxidation is a key step in the formation of volatile flavor compounds such as aldehydes, the content of these flavor compounds increases accordingly.
It should be noted that the above interpretation remains speculative based on existing literature and requires further validation through functional experiments. We have already appropriately moderated the relevant statements in the Discussion section to avoid overinterpretation as established regulatory mechanisms. We thank the reviewer again for this valuable comment, which has helped us present our findings more rigorously.
Comments 7: Minor Comments
Numerous grammatical errors and awkward phrasing are present (e.g., verb tense inconsistencies, article usage, punctuation).
“KEEG” should be corrected to “KEGG”.
Some references appear inconsistently formatted.
Figure legends require clearer statistical annotations (exact p-values instead of only * or **).
OPLS-DA validation metrics (R², Q², permutation test) should be provided.
Clarify whether raw RNA-seq data are deposited in a public repository (GEO/SRA). Current data availability statement limits reproducibility.
Strengths of the Study
Integration of lipidomics, volatilomics, and transcriptomics.
Use of GC×GC-TOF-MS for comprehensive VOC profiling.
Focus on indigenous cattle breed with breeding relevance.
Identification of candidate genes linked to fatty acid metabolism.
Response 7:Language issues: The manuscript has undergone thorough language editing to correct grammatical errors, verb tense inconsistencies, article usage, and punctuation. The overall readability has been significantly improved.
“KEEG” correction: All instances of “KEEG” have been corrected to “KEGG”.
Inconsistent references: The reference list has been reformatted to meet the journal’s requirements, ensuring consistency throughout.
Figure legends with statistical annotations: We have revised the figure legends to include exact p-values instead of using only asterisks (*) to indicate significance levels.
OPLS-DA validation metrics: We have added the missing OPLS-DA validation metrics in the Results section.
RNA-seq data availability: The samples used in this study originated from indigenous breed resources in Anhui Province. The associated data are related to ongoing breeding applications and intellectual property protection, and therefore are not suitable for public release at this stage. If deemed necessary by the reviewer or editor, we are willing to deposit the raw data in a public repository (e.g., GEO) upon manuscript acceptance, or comply with any journal-specific data deposition requirements.
Reviewer 2 Report
Comments and Suggestions for AuthorsAbstract: please indicate the values of a high fatty acid content and a low fatty acid content.
Keywords: they must be different from those used in the title; thus, please change the keywords Dabieshan cattle, fatty acid, volatile flavor compounds. This is required to increase indexing and impact of the paper.
Material and methods:
- explain why the authors used "LD between the 12 and 13th ribs of each cattle"
- Please add more information how the authors "based on the total fatty acid content of the LD, eight phenotypically extreme individuals were selected for subsequent analysis". The information must be described in a manner that other researchers can reproduce the work.
Results:
- Unit values of Figure 1A are missing
- it's not clear if the 8 groups (H and L) are from 8 individual cattles. If so, what about the remaining 12 cattle?
- Line 239-240: the relative contents of aldehydes and hydrocarbons were higher in the group H than in the L group; this comparison was made using statistics? If so, in what level of probability the samples are different?
Discussion: please verify if its possible to merge results and discussion in a single section, to provide a more flowability reading.
Author Response
Comments 1: explain why the authors used "LD between the 12 and 13th ribs of each cattle"
Response1: We thank the reviewer for the question. The Longissimus dorsi (LD) muscle between the 12th and 13th ribs was selected as the sampling site for two main reasons. First, key meat quality indicators such as marbling score and ribeye area are officially evaluated at the cross section between the 12th and 13th ribs in international beef grading systems (e.g., USDA). Using this site ensures that the measured indicators are directly aligned with industry standards. Second, this sampling location is the most commonly used in similar studies, as supported by extensive literature. This allows our results to be reliably compared with previous reports, ensuring data comparability.
Comments 2: Please add more information how the authors "based on the total fatty acid content of the LD, eight phenotypically extreme individuals were selected for subsequent analysis". The information must be described in a manner that other researchers can reproduce the work.
Response2: We thank the reviewer for this valuable comment. We have added detailed information in the Methods section. Specifically, for each of the 20 DBS cattle, we first calculated the total fatty acid content as the sum of all individual fatty acids (μg/g ) measured by GC. Then, the 20 cattle were ranked from highest to lowest total fatty acid content. The top 4 animals were assigned to the high group (H group, n=4), and the bottom 4 animals to the low group (L group, n=4). The remaining 12 animals with intermediate total fatty acid levels were excluded from subsequent transcriptomic and metabolomic analyses. This procedure is now clearly stated in the Methods section (or Results section) to ensure reproducibility. The fatty acid profiles of the H and L groups are presented in Supplementary Table S1.
Comments 3: Unit values of Figure 1A are missing
Response3: Thank you for pointing this out. The units for Figure 1A have been added in the revised figure.
Comments 4: it's not clear if the 8 groups (H and L) are from 8 individual cattles. If so, what about the remaining 12 cattle?
Response4: Yes, the H group (n=4) and L group (n=4) are from eight individual cattle. The remaining 12 cattle with intermediate total fatty acid levels were not discarded; their data are part of other ongoing studies on meat quality traits of Dabieshan cattle, and therefore cannot be disclosed in this paper. This study focuses only on the differences between the high and low fatty acid extreme phenotypes, and the 46 fatty acid profiles of the H and L groups are presented in Supplementary Table S1.
Comments 5: Line 239-240: the relative contents of aldehydes and hydrocarbons were higher in the group H than in the L group; this comparison was made using statistics? If so, in what level of probability the samples are different?
Response 5: Thank you for your question. The statement that “the relative contents of aldehydes and hydrocarbons were higher in the H group than in the L group” in lines 239–240 is based on a direct comparison of the summed values of all aldehydes and all hydrocarbons between the two groups, without statistical testing. This description simply reflects the overall trend of higher total contents in the H group and does not involve any probability level of significance.
Comments 6: please verify if its possible to merge results and discussion in a single section, to provide a more flowability reading.
Response6: Thank you for your suggestion. After careful consideration, we would prefer to keep the original structure. The Discussion section has now been sufficiently streamlined to avoid redundancy with the Results section. We appreciate your input and have made efforts to improve the coherence and readability of the text within the existing structure.
Reviewer 3 Report
Comments and Suggestions for AuthorsMS ID foods-4198020
Brief Abstract: In the MS entitled “Integrative Analysis of Volatile Flavor Compounds and Transcriptome Reveals Underlying Mechanisms Linked to Fatty Acid Content in Dabieshan Cattle,” the authors investigated the regulatory relationships among fatty acid composition, volatile flavour compounds (VOCs), and gene expression in the longissimus dorsi muscle of Dabieshan cattle. Using integrated GC×GC-TOF-MS metabolomics and transcriptomic analysis, the authors identified significant differences in VOC profiles and gene expression between the two groups of animals (high vs low fatty acid content). The high fatty acid group showed higher levels of aldehydes, esters, and hydrocarbons, whereas the low group contained more alcohols, acids, and heterocyclic compounds. Among 54 differential VOCs, (E)-2-Nonenal was identified as the key flavor compound. Transcriptomic analysis revealed 678 differentially expressed genes, eight of which were identified through functional enrichment analyses as being associated with fatty acid composition. Notably, SGPL1, KLF15 and SLC27A6 showed strong correlations with polyunsaturated fatty acids, such as C22:5n-3, C18:3n-3, C18:2n-6 and C18:1n-9c. These fatty acids were also significantly associated with key flavour compounds.
The authors conclude that the study establishes regulatory links between genes, fatty acids, and flavour compounds in Dabieshan cattle, providing new insights and molecular targets for improving flavour traits in local cattle breeds.
The manuscript covers an interesting topic and is both easy to understand and well structured. The sections on the introduction, materials and methods, results, discussion and conclusions are all well organised. I have a few minor suggestions for the authors, which are reported below:
Lines 84-88: The primary objectives were to elucidate the genetic basis of meat quality traits in DBS cattle, identify key candidate regulatory genes, and investigate how different fatty acid levels influence meat quality. The findings will provide theoretical support and molecular targets for the conservation, utilization, and genetic improvement of superior meat quality characteristics in DBS cattle.
"The aim of the study" must be clearly stated, without including results or other such information. I suggest rewriting it.
Lines 187-189: Statistical analysis was then carried out in SPSS 27.0. To evaluate intergroup differences, nonparametric tests were employed, defining p < 0.05 as statistically significant and p < 0.01 as highly significant. Data from four biological replicates per group are presented as mean ± SD (mean ± standard deviation).
This paragraph must indicate the statistical model applied (not just the SW). The correlation analysis must also be detailed.
Lines 198-206: To investigate the genetic regulatory network of fatty acid profiles and the mechanism of flavor formation in the LD of DBS cattle, we collected 20 samples with consistent genetic backgrounds and measured their fatty acid contents (Supplementary Table 1). Based on total fatty acid content, the samples were divided into H (n=4) and L (n=4) groups (Fig. 1A). To evaluate the degree of variation between and within groups, PCA was performed on all samples. As shown in Fig. 1B, the respective contributions of the first two principal components to the total variance were 75.5% and 13.2%. The results showed that a clear separation trend was observable between the two groups in the PCA score plot. Hierarchical cluster analysis confirmed that the 46 fatty acids formed two distinct clusters, which aligned with the experimental grouping (Fig. 1C).
This section appears to be a combination of materials, methods and results. Rewrite it.
Lines 147-149: The GC oven program for the primary DB-Heavy Wax column commenced at 40°C (3 min hold), increased to 50°C at 3°C/min, then to 120°C at 6°C/min, followed by a ramp to 220°C at 5°C/min, and a final hold for 15 minutes.
Check for typos in the temperature and time (°C/min).
Line 289: groups of DBS cattle..
Check punctuation.
Line 328: abundance of UFAs.
Insert the meaning of the acronym at the beginning of the text (line 378).
Lines 481-509: There is a lot of confusion between the discussion and the conclusions. Rewrite them carefully, bearing in mind that the conclusions should also provide answers to the study's objective.
Check reference no. 7.
Supplementary material: For example, the grid 'Supplement Table S9: Differential flavour compounds in the top 20 with VIP >1'. A description of the statistical parameters is necessary, as is an indication of what the p-value refers to.
Author Response
Comments 1:Lines 84-88: The primary objectives were to elucidate the genetic basis of meat quality traits in DBS cattle, identify key candidate regulatory genes, and investigate how different fatty acid levels influence meat quality. The findings will provide theoretical support and molecular targets for the conservation, utilization, and genetic improvement of superior meat quality characteristics in DBS cattle.
"The aim of the study" must be clearly stated, without including results or other such information. I suggest rewriting it.
Response 1:Thank you for your valuable suggestion. In the revised manuscript, we have rewritten the relevant sentences.
Comments 2:Lines 187-189: Statistical analysis was then carried out in SPSS 27.0. To evaluate intergroup differences, nonparametric tests were employed, defining p < 0.05 as statistically significant and p < 0.01 as highly significant. Data from four biological replicates per group are presented as mean ± SD (mean ± standard deviation).
This paragraph must indicate the statistical model applied (not just the SW). The correlation analysis must also be detailed.
Response 2:We thank the reviewer for this important methodological comment. In the revised manuscript, we have added a clear description of the statistical model applied.The corresponding description has been incorporated into the Materials and Methods section.
Comments 3:Lines 198-206: To investigate the genetic regulatory network of fatty acid profiles and the mechanism of flavor formation in the LD of DBS cattle, we collected 20 samples with consistent genetic backgrounds and measured their fatty acid contents (Supplementary Table 1). Based on total fatty acid content, the samples were divided into H (n=4) and L (n=4) groups (Fig. 1A). To evaluate the degree of variation between and within groups, PCA was performed on all samples. As shown in Fig. 1B, the respective contributions of the first two principal components to the total variance were 75.5% and 13.2%. The results showed that a clear separation trend was observable between the two groups in the PCA score plot. Hierarchical cluster analysis confirmed that the 46 fatty acids formed two distinct clusters, which aligned with the experimental grouping (Fig. 1C).
This section appears to be a combination of materials, methods and results. Rewrite it.
Response 3:We thank the reviewer for this observation. In the revised manuscript, we have rewritten the indicated paragraph to focus solely on the results, removing any content that resembled materials and methods.
Comments 4:Lines 147-149: The GC oven program for the primary DB-Heavy Wax column commenced at 40°C (3 min hold), increased to 50°C at 3°C/min, then to 120°C at 6°C/min, followed by a ramp to 220°C at 5°C/min, and a final hold for 15 minutes.
Check for typos in the temperature and time (°C/min).
Response 4:Thank you for pointing out this issue. We have carefully checked the temperature and time units in the revised manuscript and corrected the formatting.
Comments 5:Line 289: groups of DBS cattle..
Check punctuation.
Response 5:We thank the reviewer for noticing this typographical issue. The extra punctuation has been removed in the revised manuscript. The sentence now reads correctly. We appreciate the reviewer’s careful reading.
Comments 6:Line 328: abundance of UFAs.
Insert the meaning of the acronym at the beginning of the text (line 378).
Response 6:We thank the reviewer for this careful check. The acronym UFA (unsaturated fatty acids) has already been defined at its first appearance in the revised manuscript, which is now in line 45. We appreciate the reviewer’s attention to detail.
Comments 7:Lines 481-509: There is a lot of confusion between the discussion and the conclusions. Rewrite them carefully, bearing in mind that the conclusions should also provide answers to the study's objective.
Response 7:Thank you for this valuable comment. In the revised manuscript, we have carefully reorganized the Discussion and Conclusions sections and streamlined the Discussion content to eliminate any possible confusion between the two.
Comments 8:Check reference no. 7.
Response 8:We thank the reviewer for pointing out this issue. Reference 7 has been corrected and reformatted according to the journal’s requirements in the revised manuscript.
Comments 9:Supplementary material: For example, the grid 'Supplement Table S9: Differential flavour compounds in the top 20 with VIP >1'. A description of the statistical parameters is necessary, as is an indication of what the p-value refers to.
Response 9:Thank you for this important suggestion. In the revised supplementary material, we have added footnotes below each relevant table to provide a clear description of the statistical parameters and an explanation of what the p value refers to.
