Identification and Sensory Characterization of Umami Peptides During Lager Beer Fermentation

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript discusses an integrated framework, termed “computational biochemistry–receptor docking recognition molecular sensomics,” that was established to elucidate the formation of umami peptides, receptor recognition, and sensory contribution throughout the full fermentation of lager beer. The content of the article and the results are interesting, but some aspects need improvement:

1) The references in the introduction section need to be revised. Website URLs must be formatted according to the journal’s guidelines (lines 58–60)

2) Standardize the percentages so that they are formatted to the same number of decimal places throughout the manuscript. For example, 32.0% (line 52) or 86.46% (line 57).

3) When you finish a subsection, insert 12-point spacing so the text doesn’t look cramped. The M&M section where this happens is hard to read, compared to the R&D section where the spacing is included.

4) Clarify in section 2.3.9 the confidence level used in the statistical analysis.

5) The caption for Figure 3 is confusing. Perhaps using I, II, III, IV... would be better than lowercase letters.

6) I noticed that the statistical analysis was missing from some specific figures and tables. Check if it’s possible to add it.

7) To ensure the high quality of the publication, I recommend that the manuscript undergo professional language editing to refine its grammar and style.

Author Response

We sincerely thank the reviewer for the valuable comments and constructive suggestions. We have carefully revised the manuscript accordingly, and the detailed modifications and point-by-point responses are provided in the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

I consider that the topic of the manuscript falls within the scope of Foods. Overall, the manuscript is adequately structured; however, it presents several deficiencies that need to be addressed before it can be considered for publication in Foods (MDPI). The main issues that should be corrected are listed below:

1.- Lines 2-4. In my opinion, the title includes too many methodological elements (Virtual Screening, Molecular Docking, Molecular Sensomics). I suggest the authors revise the title to prioritize the scientific contribution. A possible title could be: “Identification and Sensory Characterization of Umami Peptides During Lager Beer Fermentation.”

2.- Lines 22-25. The opening statement does not clearly establish the specific problem being addressed or the scientific relevance of the study. I suggest the authors include an introductory sentence that defines the research gap. Additionally, the objective should be stated more precisely.

3.- Lines 26-27. The number of peptides is reported, but their significance is not explained. I suggest highlighting their relevance.

4.- The abstract is difficult to follow in its current form. I suggest the authors shorten and clarify it. It should ideally include an introductory statement, objective, materials and methods, results, and conclusion. Additionally, the most important findings should be emphasized.

5.- Lines 49-53. The authors are encouraged to revise the statistical section presented in the introduction.

6.- Lines 58-59. Please verify the web links. The second link leads to a “page not found.”

7.- Lines 48-59. The text focuses on general aspects of beer consumption and relevance (including market/acceptance aspects), which do not directly contribute to the scientific problem. I suggest reducing this section and clearly highlighting the scientific problem addressed in the manuscript.

8.- Lines 76-93. I suggest including a clear statement defining the research gap.

9.- Lines 98-110. Tools such as machine learning, molecular docking, and sensomics are mentioned, but their relevance and importance are not explained. I suggest adding a brief justification supported by recent literature.

10.- Lines 111-123. The objective of the study is unclear, as it is described in a broad and descriptive manner, including multiple techniques without clearly stating what is being evaluated. It also mixes objectives with methodology. I suggest reformulating it in a clear and specific manner.

11.- Line 103. “Qi et al.” should be written as “Qi et al. [ ].”

12.- Lines 126-132. The description of the beer sample is insufficient, as key fermentation conditions (yeast type, temperature, time, wort composition) are not specified. This limits reproducibility. If possible, the authors should provide a more detailed description or, if proprietary, at least a general characterization.

13.- Line 129. A storage temperature of -4 °C is reported. Please verify this value. Was the beer frozen?

14.- Line 132. Experiments were performed in triplicate (n = 3), which may be insufficient considering the variability of beer matrices. Please justify or increase the number of replicates.

15.- Lines 150-166. The LC-MS/MS analysis lacks validation parameters (LOD, LOQ, precision, accuracy). Please include this information or provide supporting references.

16.- Line 165. The use of the internal standard (PVPL) is not justified. Please explain its selection and relevance.

17.- Lines 167-173. Machine learning tools (UMPred-FRL and ProUmami) are used without reporting validation, training dataset, or performance metrics. Please provide this information or cite supporting studies.

18.- Lines 174-192. The molecular docking procedure lacks validation (e.g., redocking, RMSD, control ligands). Please include validation of the docking protocol.

19.- Lines 193-202. The sensory analysis description is limited. Panel training, selection criteria, and validation are not specified. Please provide these details.

20.- Lines 203-251. The statistical analysis is not sufficiently described (tests, significance level, software). Please clarify and justify the methods used.

21.- Lines 261-303. The paragraph is too long; please divide it into two. The description of peptides is mainly narrative and lacks statistical analysis. Please include statistical comparisons and relevant references.

22.- In section 3.1, the analysis is largely descriptive and lacks statistical support. The number of peptides is reported, but their relative abundance and significance are not discussed. Additionally, no clear criteria are provided to define relevant peptides, and validation is insufficient. This section would benefit from multivariate analysis, comparison with previous studies, and a clearer link to umami perception.

23.- In section 3.2, the analysis relies heavily on in silico tools without experimental validation. The interpretation of binding energies as direct evidence of umami activity is questionable. Additionally, peptide selection criteria are unclear and no appropriate controls are included.

24.- In section 3.3, I suggest including additional validation of docking results or clearly stating that these are in silico predictions. Comparisons with reference ligands (e.g., glutamate) should be included, and reproducibility should be described in more detail.

25.- In section 3.4, I suggest (1) strengthening the molecular dynamics analysis, particularly convergence and reproducibility; (2) including independent replicates or justifying a single simulation; (3) improving the quantitative interpretation of RMSD, RMSF, and radius of gyration; (4) discussing limitations of MM/GBSA (e.g., lack of entropy); (5) including reference systems; and (6) moderating conclusions.

26.- In section 3.5, I recommend (1) providing more detail on sensory design and validation; (2) including more robust statistical analysis; (3) clarifying data treatment; (4) improving interpretation of attributes such as umami and bitterness; and (5) better linking sensory and molecular results.

27.- In section 3.6, I suggest specifying the type of correlation analysis and significance level, describing clustering methods and metrics, and improving integration between correlation, sensory, and molecular results.

28.- In the Results section, I encourage the authors to compare their findings with existing literature and include more references, as the discussion is currently limited.

29.- Lines 850-879. The conclusion is somewhat lengthy and should be simplified. The main contribution should be clearly stated, and the novelty better emphasized.

30.- Lines 865-879. There is no clear integration between computational and sensory results. Please strengthen this connection.

31.- In the references section, DOIs should be included for all cited articles.

Author Response

We sincerely thank the reviewer for the valuable comments and constructive suggestions. We have carefully revised the manuscript accordingly, and the detailed modifications and point-by-point responses are provided in the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript presents a highly innovative and comprehensive approach to understanding the role of umami peptides in lager beer. By combining advanced analytical techniques, such as RPLC-Q-TOF-MS, with bioinformatic predictions, molecular docking, and sensory analysis, the study offers valuable insights into the flavor profile evolution during fermentation. The research is well-structured and addresses a significant gap in the brewing industry. Below are some constructive suggestions to further enhance the clarity, robustness, and overall impact of the manuscript.

• Introduction Section: The final paragraph of the introduction (lines 76-88) describes the study's strategy (RPLC-Q-TOF-MS, bioinformatic prediction, molecular docking, and sensory analysis). However, the formulation of the main objective could be more direct and assertive. Instead of merely listing the methodological steps, add a strong closing sentence. Example: "Therefore, the main objective of this study was to elucidate the structure-activity evolution of umami peptides across the full fermentation process of lager beer, providing a molecular basis for flavor-oriented process control." This makes the study's purpose crystal clear to the reader and the reviewers.

• Lines 188-192 and 218: Equations (1) and (2) present formatting errors, possibly generated during the conversion of the editing software. Review and correct the formatting of the mathematical operators in the original manuscript to avoid confusion during the reviewers' reading. Improve the manuscript equations.

2. Methodology (Methods)

• Lines 174-186 (Homology Modeling): The study uses the crystal structure of mGluR1 (PDB: 1EWK) as a template to build the 3D model of the T1R1/T1R3 subunits. Since the sequence identity between these receptors is usually relatively low, relying solely on the Ramachandran plot for validation might be considered insufficient by more demanding reviewers. Include the percentage of sequence identity and similarity between the template and the target. It is also recommended to mention whether other structural validation metrics (such as PROSA or ERRAT) were used to provide more robustness to the constructed model.

• Line 221 (Free Energy Calculation - MM/GBSA): The authors state that "Entropy changes were neglected due to high computational cost and limited accuracy". Although this is a common practice in the literature, ignoring entropy significantly overestimates the binding energy, especially for flexible ligands like peptides. This should be explicitly addressed as a study limitation in the Discussion section to demonstrate transparency to the reviewers.

3. Results and Discussion

• Table 3 / Docking Results: The reported binding free energy values are extremely negative (e.g., −77.11-77.11−77.11 kcal/mol for the ATTSI peptide). This is a direct consequence of the omission of the entropic term mentioned in the methodology. In the discussion of these results, make it clear that these values represent relative binding energies (useful for ranking affinity and comparing peptides with each other) and not absolute free energies.

• Lines 283-332 (Peptide Screening Criterion): The study adopts the criterion "detectable across at least two key stages" to filter umami peptides and remove transient noise from the process. The justification for choosing exactly "two stages" may seem empirical or arbitrary. It is recommended to add a brief analytical explanation or a bibliographic reference that supports why detection in two stages is the ideal threshold to ensure temporal reproducibility without losing important data.

• Lines 782-797 (Astringency Reduction): The discussion suggests that peptides can weaken the astringency sensation via "competitive binding and by altering boundary lubrication". Ensure it is clear whether this statement is a hypothesis based on the literature (ensuring that the references are well-connected in the paragraph) or if it was directly proven by your sensory data. If it is speculative, using terms like "our results suggest that..." or "we hypothesize that..." helps avoid criticism regarding conclusions that extrapolate the experimental data.

Author Response

We sincerely thank the reviewer for the valuable comments and constructive suggestions. We have carefully revised the manuscript accordingly, and the detailed modifications and point-by-point responses are provided in the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

I have reviewed the manuscript foods-4238762 titled “Umami Peptides Across the Full Fermentation of Lager Beer: 2 Virtual Screening, Molecular Docking, and Molecular Sensomics” by Yashuai Wu and co-authors. The study investigates the emergence and sensory role of umami peptides during lager beer fermentation—and combines peptidomics, in silico screening, receptor docking, MD/MM-GBSA, and sensory testing. Specifically, the authors tracked peptide changes across six fermentation stages of lager beer (days 0, 1, 3, 10, 17, and 29) using RPLC-Q-TOF-MS, identified candidate umami peptides with machine-learning predictors, docked these to the T1R1/T1R3 umami receptor, selected six representative peptides, and evaluated them through taste threshold determination, molecular dynamics/MM-GBSA, and sensory addition experiments. The study concludes that umami peptides accumulate dynamically during fermentation and contribute to the umami–bitterness–aftertaste axis, while also shaping the sensory balance.

The topic is interesting and relevant to for beer flavor engineering and process control. Several points needing clarification, and some suggestions to improve the manuscript are given below.

Objectives

Objectives are clearly stated but they should be stated more explicitly in the Introduction.

Novelty

The novelty of the study lies in the integration of time-resolved beer peptidomics, AI-assisted umami peptide screening, receptor docking/MD, and molecular sensomics-style validation in a brewing context.

Abstract

The abstract would benefit from addition of one sentence clarifying the practical significance for beer quality or brewing control, reduction of the number of exact numerical values in the abstract (especially docking/MD details), and softening of claims such as “elucidated” or “provided verifiable support” unless direct validation is shown in the paper.

Introduction

• This section is too long and a bit repetitive. The authors should tighten the first 2–3 paragraphs.
• Moreover, stronger discussion of prior work on umami peptides in fermented beverages, taste-active peptides in beer, and limitations of computational peptide taste prediction could be added.

Methodology

Several methodological aspects could be clarified further. Specifically:

2.1 Samples and reagents

• Fermentation batches /Replicates: The manuscript reports that routine batches were produced (n = 3). Please clarify whether these are independent brewing batches or analytical replicates from the same batch? Were all six fermentation timepoints sampled from three independent fermentation runs?
• Fermentation process details: Please add details about yeast strain, fermentation temperature profile, oxygenation, gravity change or attenuation, maturation conditions, filtration/clarification conditions.

 

2.2 Instruments and reagents

Table 1. List of reagents and instruments could be moved to the supplementary files

2.3 Experimental procedures

• Qualitative identification and semi-quantitative analysis of beer peptides

Line 165. “semi-quantitative peptide levels were calculated with the PVPL peptide as an internal standard”. Please clarify at which stage of sample preparation PVPL was added, the final concentration used, and whether recovery was monitored.

• Efficient screening of potential umami peptides using machine learning

Why a cutoff of >0.9 in both UMPred-FRL and ProUmami was used? Please explain.

MD/MM-GBSA analysis:

The authors should provide more detail on the MD/MM-GBSA analysis, particularly whether the reported results are based on single trajectories or replicate independent runs for each peptide–receptor complex, and what were the criteria used to assess convergence.

Sensory analysis:

The manuscript states a trained panel (n = 20) and 3-AFC threshold design, but the sensory analysis description needs more details. For example, how were the sensory attributes defined and anchored, were samples blind/randomized/balanced; was there duplicate or triplicate sensory presentation; were assessors screened specifically for umami sensitivity?; how was carryover managed?; was ethanol matrix or carbonation standardized in all sensory tests? serving temperature, palate cleansers ?

Statistical analysis: The statistical analysis section should specify the exact statistical models used, whether assumptions were checked, correction for multiple testing, whether correlations are Pearson or Spearman, number of replicates tested.

Results

• Section 3.1 includes long explanations about adsorption/desorption, peptide–polyphenol–metal complexes, wall interactions, autolysis-driven redistribution,

etc. This reads like a review article embedded inside Results, and part of these should be moved to Discussion section.

Discussion

• The manuscript should more clearly distinguish between what is predicted, what is experimentally supported, and what is actually demonstrated.
• The authors should acknowledge that umami perception in beer is multicomponent, not peptide-only.
• The authors should discuss more explicitly matrix effects in beer, and synergy/ antagonism with amino acids, nucleotides, polyphenols, bitterness compounds.
• Why some peptides may increase bitterness despite being “umami peptides”? Please elaborate.
• The discussion could also include a more critical evaluation of the limitations of the study. Specifically, docking and MD do not prove receptor activation, semi-quantitative peptidomics, sensory tests do not fully reproduce endogenous matrix interactions, especially synergism with glutamate, nucleotides, and bitterness compounds; potential batch/matrix dependence (only one lager system / one production context); findings may not transfer directly to other lager styles or breweries.

Conclusions

• The conclusions need to be more conservative. Some claims appear stronger than the data support. I suggest avoiding terms such as “established” or “elucidated”
• The conclusion would be stronger if it ended with a realistic practical implication, such as the potential use of this workflow for fermentation monitoring or flavor optimization in brewing.

Author Response

We sincerely thank the reviewer for the valuable comments and constructive suggestions. We have carefully revised the manuscript accordingly, and the detailed modifications and point-by-point responses are provided in the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the careful revision of your manuscript entitled “Umami Peptides Across the Full Fermentation of Lager Beer: Virtual Screening, Molecular Docking, and Molecular Sensomics.” I would like to commend you for your thorough work and for the effort made to address the previously raised comments and suggestions, which has resulted in a significant improvement in the manuscript in terms of clarity, structure, and scientific focus. Overall, you have responded adequately to the majority of the comments, leading to a clear enhancement in the quality, coherence, and scientific rigor of the work. In my opinion, the manuscript now meets the quality standards required for publication in Foods. At this stage, I have no further comments.

Author Response

We would like to express our sincere gratitude to the reviewers for their careful evaluation of our manuscript and for the time and effort they devoted to this work. We are especially grateful for their positive comments and recognition of the value of our study. Their constructive suggestions have been highly valuable and have helped us improve the clarity, rigor, and overall quality of the manuscript. We truly appreciate the reviewers’ professionalism, insightful feedback, and thoughtful assessment, which have contributed greatly to strengthening this work. We also thank the editor for the opportunity to revise and improve the manuscript.

Reviewer 4 Report

Comments and Suggestions for Authors

The authors have clearly made a serious effort, but there are still important methodological, interpretive, and wording issues that should be corrected.

Abstract

The abstract has improved, but it remains too dense and still overstates what the data support. Please further reduce methodological and numerical detail (especially peptide counts and docking/MD-specific information) and use more conservative language throughout.

Introduction

The Introduction has improved, but it remains longer than necessary and still contains some review-style material that could be tightened substantially. In particular, the market-oriented background and the extended description of prediction tools should be shortened.

Methodology

The replication statement is clearer now. However, please clarify how these three independent batches were used in the peptidomics and single-addition sensory analyses (i.e., whether all batches were analyzed independently or pooled). Please also clarify whether peptides were added to each independent batch separately, to pooled material, or to one representative beer per stage.

Table 1. The response is noted. I leave placement of Table 1 to the editor and journal style.

Internal standard PVPL: The clarification is helpful, but the reported apparent recovery range (87.92–137.76%) indicates substantial analytical variability and likely matrix effects. Please explicitly acknowledge in the manuscript that PVPL is a surrogate rather than isotopically matched internal standard and that the resulting peptide abundance estimates should be interpreted as semi-quantitative only.

>0.9 cutoff: The rationale for using a >0.9 cutoff in both UMPred-FRL and ProUmami is now clearer and is scientifically reasonable. However, this explanation is still not stated explicitly enough in the revised manuscript itself. Please add a brief sentence in the Materials and Methods.

MD / MM-GBSA: The MD/MM-GBSA section is improved, but it still does not fully answer the key methodological concern whether each peptide–receptor complex was analyzed using a single production trajectory or replicate independent MD runs. Have the authors performed replicate independent simulations? If not, state this explicitly and acknowledge it as a limitation.

In addition, the basis for defining the 90–100 ns interval as “converged” should be justified more explicitly, rather than simply selecting a terminal trajectory window.

Sensory analysis: The sensory section is much improved. However, please clarify how “umami sensitivity” was assessed during panel training, describe the 3-AFC threshold method more clearly, and explain how fatigue was managed if samples were evaluated in triplicate.

Yeast strain: the manuscript is framed throughout as a study of lager beer, but the authors now report using Angel Yeast CS31 Belgian ale yeast (Saccharomyces cerevisiae). This appears inconsistent with a typical lager fermentation system and needs to be clarified carefully. If an ale strain was used, the terminology and scope of the manuscript may need to be adjusted.

Conclusions

The Conclusions section is still too strong and should be revised more conservatively to align with the improved Discussion.

English language

The manuscript would benefit from an additional round of professional English editing to improve readability and precision of expression.

Author Response

We sincerely thank the reviewer for the valuable comments and constructive suggestions. We have carefully revised the manuscript accordingly, and the detailed modifications and point-by-point responses are provided in the attachment.

Author Response File: Author Response.pdf