Barley Wine in Focus: NMR Metabolomics Reveals Style and Barrel Aging Differences
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe proposed article concerns the field of application of NMR metabolomics in the field of analyzing barley wine. I think that this paper can be accepted after correction of some issues. These are as follows:
Lines 122-123: “deuterated phosphate buffer solution (pD 4.4, prepared in Dâ‚‚O) containing 0.1 % TSP”. Please elaborate what phosphates (sodium, potassium, orthophosphate, hydrophosphate or dihydrophosphate) exactly were used, in which quantities and quality of the reagents.
Lines 139-140: “A comprehensive approach was used to assign and confirm metabolites. This approach integrated reference database matching (Human Metabolome Database and Bio-140 logical Magnetic Resonance Data Bank), literature comparison”. Please add the references to the literature that have been used for metabolite identification as well as to the HMDB (available on the database website in the “citing” section).
Line 155: “CHAID decision tree”. Please expand this abbreviation at the moment it is first mentioned.
Author Response
We thank the reviewer for the constructive and helpful comments. All requested clarifications have been implemented in the revised manuscript as detailed below.
Comment 1: Lines 122-123: “deuterated phosphate buffer solution (pD 4.4, prepared in Dâ‚‚O) containing 0.1 % TSP”. Please elaborate what phosphates (sodium, potassium, orthophosphate, hydrophosphate or dihydrophosphate) exactly were used, in which quantities and quality of the reagents.
Response 1: We thank the reviewer for this helpful comment. The manuscript has been revised to specify the identity and amount of each component of the phosphate buffer (NaHâ‚‚POâ‚„, TSP, NaN₃, H₃POâ‚„) and to indicate that all reagents were analytical grade. The text in lines 123–127 has been revised and now reads as follows:
“Then, 500 µL of the degassed sample were combined with 50 µL of a 1.0 M deuterated phosphate buffer solution (pD 4.4, prepared in Dâ‚‚O) containing 1.27 g NaHâ‚‚POâ‚„ per 10 mL Dâ‚‚O, 0.1 % TSP (sodium salt of 3-(trimethylsilyl)-2,2,3,3-tetradeuteropropionic acid) as an internal chemical shift reference, 0.05 % NaN₃ as a preservative and 55 µL H₃POâ‚„. All reagents were of analytical grade.”
Comment 2: Lines 139-140: “A comprehensive approach was used to assign and confirm metabolites. This approach integrated reference database matching (Human Metabolome Database and Bio-140 logical Magnetic Resonance Data Bank), literature comparison”. Please add the references to the literature that have been used for metabolite identification as well as to the HMDB (available on the database website in the “citing” section).
Response 2: Thank you for pointing this out. We have now included the appropriate references for the HMDB, BMRB, and peer-reviewed NMR literature used for metabolite assignment. The corresponding citations have been included in the revised manuscript. Specifically, the revised manuscript now cites:
- HMDB: Wishart, D.S.; Guo, A.; Oler, E.; Wang, F.; Anjum, A.; Peters, H.; Dizon, R.; Sayeeda, Z.; Tian, S.; Lee, B.L.; et al. HMDB 5.0: the Human Metabolome Database for 2022. Nucleic Acids Res. 2022, 50, D622–D631. https://doi.org/10.1093/nar/gkab1062
- BMRB: Ulrich, E.L.; Akutsu, H.; Doreleijers, J.F.; Harano, Y.; Ioannidis, Y.E.; Lin, J.; Livny, M.; Mading, S.; Maziuk, D.; Miller, Z.; et al. BioMagResBank. Nucleic Acids Res. 2008, 36, D402–D408. https://doi.org/10.1093/nar/gkm957
- Beer NMR metabolite assignment studies: Chorbadzhiev, P.; Gerginova, D.; Simova, S., Weiss or Wit: Chemical Profiling of Wheat Beers via NMR-Based Metabolomics. Foods 2025, 14, 1621. https://doi.org/10.3390/foods14091621 ; Almeida, C.; Duarte, I.F.; Barros, A.S.; Rodrigues, J.; Spraul, M.; Gil, A.M. Composition of beer by ¹H NMR spectroscopy: Effects of brewing site and date of production. J. Agric. Food Chem. 2006, 54, 700–706. https://doi.org/10.1021/jf0526947
Comment 3: Line 155: “CHAID decision tree”. Please expand this abbreviation at the moment it is first mentioned.
Response 3: We agree and have revised the manuscript accordingly. The abbreviation is now expanded upon first occurrence. The updated text reads:
“The following methods were employed: hierarchical cluster analysis (HCA), orthogonal partial least squares discriminant analysis (OPLS-DA), orthogonal two-way partial least squares discriminant analysis (O2PLS-DA) and Chi-square automatic interaction detector (CHAID) decision tree.“
Reviewer 2 Report
Comments and Suggestions for AuthorsThis paper is focused on the characterization of Barley wine beers, using NMR matabolomics strategy. This is the first study to characterize in a sysmetamic way Barlley wines using this analytic approach.
A total of 20 Barley wine beers have been analyzed. Perhaps this number of samples may be a bit small. It would have been better to have more samples, but it’s still not bad overall.
As a result of this research, 55 metabolites have been identified and quantified , and only 5 were considered as markers in order to classify the samples according to their style and barrel aging time. These compounds are: 5-hydroxymethylfurfural (HMF), acetaldehyde, mammose and tryptophan.
In my opinion, this paper shows the characterization of Barley wine beers in a very complete way, using different statistical approaches. The results are presented very clearly through the various graphs, allowing for the formulation of scientifically valuable conclusions.
Author Response
Comment 1: This paper is focused on the characterization of Barley wine beers, using NMR matabolomics strategy. This is the first study to characterize in a sysmetamic way Barlley wines using this analytic approach.
A total of 20 Barley wine beers have been analyzed. Perhaps this number of samples may be a bit small. It would have been better to have more samples, but it’s still not bad overall.
As a result of this research, 55 metabolites have been identified and quantified , and only 5 were considered as markers in order to classify the samples according to their style and barrel aging time. These compounds are: 5-hydroxymethylfurfural (HMF), acetaldehyde, mammose and tryptophan.
In my opinion, this paper shows the characterization of Barley wine beers in a very complete way, using different statistical approaches. The results are presented very clearly through the various graphs, allowing for the formulation of scientifically valuable conclusions.
Response 1: We sincerely thank Reviewer 2 for their positive and encouraging comments on our work. We appreciate the recognition of the novelty and clarity of our study, and we are grateful for the constructive feedback provided.
Reviewer 3 Report
Comments and Suggestions for AuthorsThe article is very interesting and also the subject, but some correction in the bibliography are needed:
in general it has to add references because maximum we have one reference for subject and is not enough. Otherwise there are sentences I found without reference at all:
-In Europe, brewers often use casks that previously held whisky, many of these casks were originally ex-bourbon or ex-sherry barrels.
-Recent analytical research has expanded our understanding of how wood and aging affect beer chemistry. Gas chromatography-mass spectrometry (GC-MS) and liquid chromatography mass spectrometry (LC-MS) are commonly used to profile the volatile and 80 phenolic compounds extracted from oak during maturation.
-Similarly, LC-based studies have demonstrated that the origin of the oak and the malt composition significantly impact the phenolic and bioactive profiles of wood-aged beers.
For what concern the conclusion it can be added the expression
"Also if farther analyses are needed to confirm our findings" because this can be a preliminary article because with only 20 samples is not possible to state robust scientific method.
Author Response
We thank the reviewer for the useful comments concerning the literature support and the conclusion. These points have now been corrected as outlined below
Comment 1: Some correction in the bibliography are needed:
in general it has to add references because maximum we have one reference for subject and is not enough. Otherwise there are sentences I found without reference at all:
-In Europe, brewers often use casks that previously held whisky, many of these casks were originally ex-bourbon or ex-sherry barrels.
-Recent analytical research has expanded our understanding of how wood and aging affect beer chemistry. Gas chromatography-mass spectrometry (GC-MS) and liquid chromatography mass spectrometry (LC-MS) are commonly used to profile the volatile and 80 phenolic compounds extracted from oak during maturation.
-Similarly, LC-based studies have demonstrated that the origin of the oak and the malt composition significantly impact the phenolic and bioactive profiles of wood-aged beers.
Response 1: We thank the reviewer for highlighting these omissions. The introduction has now been revised and expanded with multiple peer-reviewed references supporting each statement. In particular, we have added:
- On the use of ex-bourbon and ex-sherry whisky casks in European brewing:
Hornsey, I. A History of Beer and Brewing; Royal Society of Chemistry: Cambridge, UK, 2003. https://doi.org/10.1039/9781847550026 - On recent analytical studies (GC–MS and LC–MS) addressing wood- and barrel-aged beers:
Pang, X.; Yin, H.; Li, J.; Shi, Y.; Yang, Z. Molecular insights into the contribution of oak barrel aging to the aroma of beer with high alcohol content using SAFE-GC-O/AEDA and OAV calculation. Food Chem. 2025, 491, 145329. https://doi.org/10.1016/j.foodchem.2025.145329 - On LC-based studies showing the influence of oak origin and malt composition:
Machado Jr, J.C.; Nicola, P.D.; Viegas, O.; Santos, M.C.; Faria, M.A.; Ferreira, I.M. Bioactive properties and phenolic composition of wood-aged beers: Influence of oak origin and the use of pale and dark malts. Foods 2023, 12, 1237. https://doi.org/10.3390/foods12061237;
Tarko T.; Krankowski, F.; Duda-Chodak, A. The impact of compounds extracted from wood on the quality of alcoholic beverages. Molecules 2023, 28, 620. https://doi.org/10.3390/molecules28020620
Comment 2: For what concern the conclusion it can be added the expression
"Also if farther analyses are needed to confirm our findings" because this can be a preliminary article because with only 20 samples is not possible to state robust scientific method.
Response 2: We agree with the reviewer that the small sample size is a limitation. We have included an explicit statement in the conclusion that acknowledges this limitation and emphasizes the preliminary nature of the dataset. This sentence is now added to Conclusion:
“While the dataset clearly shows molecular trends, further analysis using larger, more controlled samples is necessary to confirm and expand these findings.”
Reviewer 4 Report
Comments and Suggestions for AuthorsThe manuscript presents an original and valuable contribution to the field of brewing chemistry and metabolomics. The study is methodologically sound, clearly structured, and written in a scientific yet engaging style.
The study presents a high level of novelty and sciefintific relevance, as it constitutes the interesting systematic NMR metabolomic investigation of barley wines and their cereal-based analogues. The interpretation of datai is clear and well supported by statistical evidence, particularly in the differentiation between various cereal types, sub-styles, and barrel-aging treatments. Moreover, the practical implications of this work are significant, as the findings offer valuable potential applications in the authentication of beers and in quality control processes within the brewing industry.
My detailed comments:
The study includes 20 commercial samples with diverse origins. Please discuss in more depth how yeast strain, and fermentation variability might influence the observed metabolomic differences and the robustness of the statistical models
The manuscript could be strengthened by discussing potential sensory implications of key discriminant metabolites (e.g., fusel alcohols, esters, furanic compounds). Even a theoretical mapping between metabolite classes and sensory attributes would improve the practical interpretation.
Figures 4–6 are highly informative but visually dense. Please consider improving label size, simplifying legends, and increasing color contrast for better clarity.
Ensure consistent use of units (e.g., mg/L vs. g/L) throughout the text.
Verify reference formatting according to MDPI style (DOIs and URLs are missing full hyperlink format).
Author Response
We thank the reviewer for the positive and encouraging evaluation of our manuscript. We are grateful for the acknowledgment of the study's novelty, methodological rigor, and practical relevance. We have addressed the specific comments below.
Comment 1: The study includes 20 commercial samples with diverse origins. Please discuss in more depth how yeast strain, and fermentation variability might influence the observed metabolomic differences and the robustness of the statistical models.
Response 1: We thank the reviewer for this valuable suggestion. We have now added a short discussion addressing the potential influence of yeast strain, fermentation temperature, and process variability on the observed chemical differences. This discussion has been incorporated into Section 3.4, where we note that samples inherently carry variability in yeast metabolism and fermentation conditions, and we clarify how OPLS-DA/O2PLS-DA and CHAID account for such variation through supervised and orthogonal components.
“Because the beers analyzed in this study were produced by different craft breweries, variations in yeast strain, fermentation temperature, wort composition, and pitching rate were expected, and these variations may have contributed to metabolite variability. Yeast-dependent traits, such as fusel alcohol formation via the Ehrlich pathway, ester production, and organic acid assimilation, influence how samples are positioned in multivariate space. While this process-level heterogeneity introduces background noise, supervised models such as OPLS-DA and O2PLS-DA can effectively separate the systematic variation associated with cereal type, substyle, and barrel aging from brewery- or fermentation-specific variation. Similarly, the CHAID decision tree partitions samples based on the metabolites that discriminate the strongest, thus mitigating the influence of such variability. The consistency of the key discriminant metabolites across groups indicates that these markers remain robust despite differences in craft brewing practices.”
Comment 2: The manuscript could be strengthened by discussing potential sensory implications of key discriminant metabolites (e.g., fusel alcohols, esters, furanic compounds). Even a theoretical mapping between metabolite classes and sensory attributes would improve the practical interpretation.
Response 2: We thank the reviewer for this valuable suggestion. The sensory relevance of key discriminant metabolites is already addressed throughout the Results and Discussion section. In Sections 3.1, 3.2, 3.3, and 3.4, we describe how metabolites such as higher alcohols, esters, furanic compounds, and complex sugars relate to sweetness, viscosity, mouthfeel, fruity esters, fusel warmth, and caramelized/oxidative notes. These explanations already link the chemical markers to their expected sensory contributions. As this interpretation is already integrated into the manuscript, no additional changes were required.
Comment 3: Figures 4–6 are highly informative but visually dense. Please consider improving label size, simplifying legends, and increasing color contrast for better clarity.
Response 3: We appreciate the reviewer’s comments. After careful evaluation, we have decided to retain the current figure design. Figures 4-6 follow the standard formatting of chemometric graphics used in Beverages and other MDPI journals, with the correct minimum font size, color palette, and contrast level required by the publisher. The other three reviewers did not report any readability concerns. We believe the current layout provides a clear and accurate visual representation of the multivariate models without compromising interpretability. For these reasons, we respectfully maintain the original figures.
Comment 4: Ensure consistent use of units (e.g., mg/L vs. g/L) throughout the text.
Response 4: We thank the reviewer for this observation. Concentrations are indeed expressed in mg/L throughout the manuscript, except in cases where metabolite levels exceed several thousand mg/L (e.g., ethanol, glycerol, maltodextrin, glucose, fructose, sucrose). In these cases, the use of g/L avoids extremely large numerical values and follows standard conventions in analytical chemistry and beer metabolomics. We have rechecked the manuscript to ensure that units are used consistently and intentionally.
Comment 5: Verify reference formatting according to MDPI style (DOIs and URLs are missing full hyperlink format).
Response 5: The references have been fully updated to match MDPI formatting, including DOI hyperlinks where applicable. Missing DOIs have been added and URL formatting corrected.
