Comparative Study of Key Aroma Components in Rice of Different Aroma Types Using Flavor Metabolomics

Round 1

Reviewer 1 Report (Previous Reviewer 1)

Comments and Suggestions for Authors

The identification of compounds is unreliable - manuscript should be rejected.
How were the isomers identified? For example: (2R,4aS,5R,8aS)
-2,5-Diallyl-1-methyldecahydroquinoline.
Are phthalates produced by plants?

Author Response

Comments 1： The identification of compounds is unreliable - manuscript should be rejected.
How were the isomers identified? For example: (2R,4aS,5R,8aS)-2,5-Diallyl-1-methyldecahydroquinoline. Are phthalates produced by plants?

Response 1：Thank you very much to the judges for your valuable comments and suggestions. Based on your suggestions, all the volatile compounds identified in the article were re-examined and all of them were compared with previous studies. A total of 8 major categories and 75 types of volatile flavor substances were identified, and the relevant parts of the article were adjusted. Please refer to the newly submitted version for the specific adjustments. Once again, we would like to express our gratitude to the teacher for the comments and suggestions given.

Reviewer 2 Report (Previous Reviewer 2)

Comments and Suggestions for Authors

Comparative study of key aroma components in rice of different aroma types using flavor metabolomics

Qi, H. Ren，H. Yang, and M. Zhang

The manuscript, entitled “Comparative study of key aroma components in rice of different aroma types using flavor metabolomics”, by Qi et al., is an article that provides insight into the flavor differences between 11 types of rice.

The study presents an interesting concept, and the authors have addressed all my previous comments effectively. However, I have a few additional observations:

1. In Tables 3 and 4, a symbol such as () is used next to “Mean content (μg/L)”, e.g., Mean content (μg/L)*, to indicate that the values are based on a semi-quantitative analysis using the area of 2-methyl-3-heptanone (IS).
2. Please, increase also the font size in all figures to enhance readability.

Author Response

Comments 1： In Tables 3 and 4, a symbol such as () is used next to “Mean content (μg/L)”, e.g., Mean content (μg/L)*, to indicate that the values are based on a semi-quantitative analysis using the area of 2-methyl-3-heptanone (IS).

Response 1：In Tables 3 and 4, a symbol such as (³The values are based on a semi-quantitative analysis using the area of 2-methyl-3-heptanone (IS)) is used to indicate that the values are based on a semi-quantitative analysis using the area of 2-methyl-3-heptanone (IS).

Comments 2： Please, increase also the font size in all figures to enhance readability.

Response 2：The font size in the chart has been adjusted.

Reviewer 3 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

The authors have addressed most of the suggestions and questions raised in the previous manuscript. Only a couple of comments.
Fig. 3. The letters in the different graphs are usually included in the non-appropriate rectangular bordering the figure.
Figure 4. Legends inside the rectangles remain illegible. 

Author Response

Comments 1： Fig. 3. The letters in the different graphs are usually included in the non-appropriate rectangular bordering the figure.

Response 1：The letters in different charts have been adjusted to the appropriate positions.

Comments 2： Figure 4. Legends inside the rectangles remain illegible. 

Response 2：The picture has been made clearer.

Reviewer 4 Report (New Reviewer)

Comments and Suggestions for Authors

The manuscript presents a comprehensive comparative analysis of volatile aroma compounds in rice samples with different aroma characteristics, using sensory evaluation combined with HS-SPME-GC-MS and GC-IMS. The topic is relevant to food chemistry, cereal science, and flavor metabolomics, and the integration of sensory data with instrumental analysis is a notable strength. The study provides valuable insights into the aroma classification of rice varieties commonly available in China.

However, several issues must be addressed before the manuscript can be considered for publication. My detailed comments are listed below.

Lines 11–15: The abstract is generally well structured; however, the phrase “molecular basis for the flavor differences” is somewhat overstated, as the study is based on volatile profiling rather than molecular or genetic mechanisms.

Lines 19–21: The abstract reports 291 volatile compounds, whereas later sections (Results 3.2.1) report 130 compounds identified by GC-MS. This inconsistency must be corrected.

Lines 35–37: The citation format is incorrect (“[1.2]” and “[3.4]”). Please correct it to the standard format (“[1,2]”, “[3,4]”).

Line 115: “teen aroma attributes” should be corrected to “ten aroma attributes.”

Lines 125–129: Sensory evaluation was performed on raw rice. This should be explicitly justified or acknowledged as a limitation, since rice is typically consumed after cooking.

Lines 217–219: Cluster analysis is presented without validation . Please justify the clustering approach.

Table 2 / Section 2.1: Chemicals mentioned in Table 2 should also be listed in subsection 2.1 (Chemicals and reagents).

Line 97: Was only one commercial sample per variety purchased? If so, were these samples analyzed only through triplicate instrumental measurements? Please clarify the sampling design.

Lines 145–146: The sentence “Headspace injection conditions included incubation at 60°C, preheating for 15 min, adsorbing for 30 min, and desorption for 4 min” is more appropriate for subsection 2.4.1, as it describes the extraction procedure on the SPME fiber rather than GC-MS analysis.

Line 246 / Table 3 is extremely long and difficult to read. Consider moving the complete compound list to the Supplementary Materials and including in the main manuscript a reduced table containing only VOCs with ROAV ≥ 1. I also recommend removing some columns (e.g., molecular formula and CAS number) and adding retention time alongside RI.

The category “Organic oxygen compounds” is too broad; more specific chemical classes should be used (e.g., alcohols, aldehydes, ketones, esters). Compound names should follow IUPAC nomenclature, but commonly used trivial names may be added for better readability. For example:

• Compound 72 ((Z)-2-Furanmethanol, 5-ethenyltetrahydro-α,α,5-trimethyl-) is commonly known as cis-linalool oxide.

• Compound 68 ((Z)-Cyclohexanol, 1-methyl-4-(1-methylethyl)-) is commonly known as p-menthan-1-ol.

Additionally, several compounds may originate from contamination. Compounds 36, 101, 110, 117, 125, and 130 are not considered natural rice volatiles, and their presence in Table 3 is confusing. The authors should carefully examine for possible contamination or misidentification. Given the number of questionable compounds, the reliability of compound identification in Table 3 is a serious concern. I strongly recommend re-evaluating all identified compounds, verifying their structures, and checking the literature to confirm whether they have been previously reported as rice VOCs.

Based on these comments, I recommend major revision.

Author Response

Comments 1： Lines 11–15: The abstract is generally well structured; however, the phrase “molecular basis for the flavor differences” is somewhat overstated, as the study is based on volatile profiling rather than molecular or genetic mechanisms.

Response 1：The molecular basis has been adjusted to the material basis.

Comments 2： Lines 19–21: The abstract reports 291 volatile compounds, whereas later sections (Results 3.2.1) report 130 compounds identified by GC-MS. This inconsistency must be corrected.

Response 2：Consistency adjustments have been made throughout the entire text.

Comments 3： Lines 35–37: The citation format is incorrect (“[1.2]” and “[3.4]”). Please correct it to the standard format (“[1,2]”, “[3,4]”).

Response 3：The citation format has been adjusted.

Comments 4： Line 115: “teen aroma attributes” should be corrected to “ten aroma attributes.”

Response 4：The spelling errors have been corrected.

Comments 5： Lines 125–129: Sensory evaluation was performed on raw rice. This should be explicitly justified or acknowledged as a limitation, since rice is typically consumed after cooking.

Response 5：Taking into account the significant influence of the aroma of raw rice on consumers' purchasing decisions, this study focuses on raw rice as the research object, aiming to verify the feasibility of rice aroma differentiation based on the main varieties available in China.The above statement has been supplemented in the text.

Comments 6： Lines 217–219: Cluster analysis is presented without validation . Please justify the clustering approach.

Response 6：The clusters was evaluated by the elbow method using the Within-Cluster Sum of Squares (WCSS). The relevant content has been added to the article.

Comments 7： Table 2 / Section 2.1: Chemicals mentioned in Table 2 should also be listed in subsection 2.1 (Chemicals and reagents).

Response 7：The chemical substances mentioned in Table 2 have been listed in Section 2.1 (Chemical Substances and Reagents).

Comments 8： Line 97: Was only one commercial sample per variety purchased? If so, were these samples analyzed only through triplicate instrumental measurements? Please clarify the sampling design.

Response 8：Only one commercial sample was purchased. Prior to the experiments, each sample was divided into three portions for subsequent research. These samples were vacuum-packed and stored at 4°C. The relevant content has been added to the article.

Comments 9： Lines 145–146: The sentence “Headspace injection conditions included incubation at 60°C, preheating for 15 min, adsorbing for 30 min, and desorption for 4 min” is more appropriate for subsection 2.4.1, as it describes the extraction procedure on the SPME fiber rather than GC-MS analysis.

Response 9：This part has been adjusted.

Comments 10： Line 246 / Table 3 is extremely long and difficult to read. Consider moving the complete compound list to the Supplementary Materials and including in the main manuscript a reduced table containing only VOCs with ROAV ≥ 1. I also recommend removing some columns (e.g., molecular formula and CAS number) and adding retention time alongside RI.

Response 10：After re-adjusting the volatile compounds, this part has become much easier to read.

Comments 11： The category “Organic oxygen compounds” is too broad; more specific chemical classes should be used (e.g., alcohols, aldehydes, ketones, esters). Compound names should follow IUPAC nomenclature, but commonly used trivial names may be added for better readability. For example:

Compound 72 ((Z)-2-Furanmethanol, 5-ethenyltetrahydro-α,α,5-trimethyl-) is commonly known as cis-linalool oxide.

Compound 68 ((Z)-Cyclohexanol, 1-methyl-4-(1-methylethyl)-) is commonly known as p-menthan-1-ol.

Response 11：The classification has been adjusted to include alcohols, aldehydes, ketones, esters, etc. The names of the compounds have been adjusted in accordance with the IUPAC nomenclature rules.

Comments 12： Additionally, several compounds may originate from contamination. Compounds 36, 101, 110, 117, 125, and 130 are not considered natural rice volatiles, and their presence in Table 3 is confusing. The authors should carefully examine for possible contamination or misidentification. Given the number of questionable compounds, the reliability of compound identification in Table 3 is a serious concern. I strongly recommend re-evaluating all identified compounds, verifying their structures, and checking the literature to confirm whether they have been previously reported as rice VOCs.

Compound 68 ((Z)-Cyclohexanol, 1-methyl-4-(1-methylethyl)-) is commonly known as p-menthan-1-ol.

Response 12：Thank you very much to the judges for your valuable comments and suggestions. Based on your suggestions, all the volatile compounds identified in the article were re-examined and all of them were compared with previous studies. A total of 8 major categories and 75 types of volatile flavor substances were identified, and the relevant parts of the article were adjusted. Please refer to the newly submitted version for the specific adjustments. Once again, we would like to express our gratitude to the teacher for the comments and suggestions given.

Round 2

Reviewer 1 Report (Previous Reviewer 1)

Comments and Suggestions for Authors

Identification is unreliable - e.g. 2,4,7,9-Tetramethyl-5-decyn-4,7-diol and many other compounds. What is proof of identification?

Author Response

Comments 1： Identification is unreliable - e.g. 2,4,7,9-Tetramethyl-5-decyn-4,7-diol and many other compounds. What is proof of identification?

Response 1：Thank you very much to the judges for your valuable comments and suggestions. Based on your suggestions, all the volatile compounds identified in the article were re-examined and all of them were compared with previous studies. In previous studies, three articles mentioned the detection of the substance 2,4,7,9-Tetramethyl-5-decyn-4,7-diol in rice. The relevant articles are shown in the attachment. These articles indicate that 2,4,7,9-tetramethyl-5-decyn-4,7-diol may be a putative contaminants, but still classify it as a volatile compound found in rice. Therefore, this study also included it in the list. However, considering the teacher's advice, in order not to mislead the subsequent readers, the relevant content has been adjusted. Please refer to the newly submitted version for the specific adjustments. Once again, I would like to express my gratitude to the teacher for the meticulous guidance provided regarding the relevant content in the article.

Author Response File: Author Response.pdf

Reviewer 4 Report (New Reviewer)

Comments and Suggestions for Authors

The authors have implemented the requested revisions in the manuscript. I suggest proceeding with the publication process.

Author Response

Comments 1： The authors have implemented the requested revisions in the manuscript. I suggest proceeding with the publication process.

Response 1：Thank you very much to the judges for your valuable comments and suggestions.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Comparative study of key aroma components in rice of different aroma types using flavor metabolomics

Below are a few comments.

1. The target compound was qualitatively analyzed by searching and comparing the GC RI (NIST 2020) database and the IMS migration time database built into the VOCal software - What is the reliability of the identification?
2. Are isomers identified by the library? For example: 1,4-Cyclohexanediamine, cis-
3. For example: Oxirane, 3-ethyl-2,2-dimethyl- Does this compound occur naturally?
4. 3-Methylsalicylic acid, 2TMS derivative - This is an artefact originating from the derivatisation process.
5. Stearic acid, TBDMS derivative - This is an artefact originating from the derivatisation process.
6. Silanol, trimethyl- from the GC column.
7. The percentage of volatile compounds in rice with different aromas was calculated by dividing the individual peak area of VOCs by the total peak area of all VOCs. - by the Internal normalisation method? Relative quantities wee calculated?

The results are unbelievable - Identification of contaminants.

Author Response

Comments 1： The target compound was qualitatively analyzed by searching and comparing the GC RI (NIST 2020) database and the IMS migration time database built into the VOCal software - What is the reliability of the identification?

Response 1：After the measurement, the mass spectrum of the volatile compounds were then compared with data from the NIST 2017 database. Then, the actual retention index (RI) value is calculated according to the relevant research methods of the same laboratory. The RI were compared with the values reported in the literature. When the difference is <50, it was inferred that the two match. This is also the common approach for qualitative analysis at present. Referring to the research of Ye. [Guodong, Ye; Lina, Guan; Min, Zhang; Sixuan, Li; Yongjie, Mi. Study on the correlation between aroma compounds and texture of cooked rice: A case study of 15 Japonica rice species from Northeast China. Cereal Chemistry. 2025,102:167–180]

Comments 2： Are isomers identified by the library? For example: 1,4-Cyclohexanediamine, cis-

Comments 3： For example: Oxirane, 3-ethyl-2,2-dimethyl- Does this compound occur naturally?

Comments 4： 3-Methylsalicylic acid, 2TMS derivative - This is an artefact originating from the derivatisation process.

Comments 5： Stearic acid, TBDMS derivative - This is an artefact originating from the derivatisation process.

Comments 6： Silanol, trimethyl- from the GC column.

Response 2-6：1,4-Cyclohexanediamine, cis- could be identified by NIST. [https://webbook.nist.gov/cgi/cbook.cgi?ID=15827-56-2]

The identified flavor compounds contain many unnecessary substances. To better identify the flavor substances in rice, the volatile substances were re-identified. For detailed information, please refer to the latest submission version. However, the identified flavor substances do not affect the subsequent analysis of the differentiated flavor substances.

Comments 7： The percentage of volatile compounds in rice with different aromas was calculated by dividing the individual peak area of VOCs by the total peak area of all VOCs. - by the Internal normalisation method? Relative quantities were calculated?

Response 7：During GC-MS analysis process, the volatile compounds were quantified (semi-quantitative analysis) by dividing the peak areas of the compounds of interest by the peak area of 2-methyl-3-heptanone as internal standard. During GC-IMS analysis process, the intensities of the volatile compounds were analyzed according to the peak volumes of the selected signal peaks by Gallery Plot analysis (v.1.0.7, G.A.S.). Visual and quantitative comparisons of volatility differences between samples were performed using the Gallery plot plug-in. Referring to the research of Liu. [Liu, D.Y.; Bai, L; Feng, X.; Chen, Y.P.; Zhang D.; Yao W.S. Characterization of Jinhua ham aroma profiles in specific to aging time by gas chromatography-ion mobility spectrometry (GC-IMS). Meat Sci. 2020, 168]

Reviewer 2 Report

Comments and Suggestions for Authors

Comparative study of key aroma components in rice of different aroma types using flavor metabolomics

1. Qi, H. REN，H. YANG, and M. ZHANG

The manuscript, entitled “Comparative study of key aroma components in rice of different aroma types using flavor metabolomics”, by Qi et al., is an article that provides insight into the flavor differences between 11 types of rice.

While the study is conceptually interesting, the authors base it entirely on the sensory evaluation conducted by a panel. It is not mentioned or demonstrated that this panel had professional training, as in the case of cacao, liquor, tea, etc., nor is the questionnaire used to qualify the taste provided.

Furthermore, the manuscript currently lacks essential data and fails to adhere to the "Proposed Minimum Reporting Standards for Chemical Analysis" by the Chemical Analysis Working Group (CAWG) of the Metabolomics Standards Initiative (MSI)[1]. It should include a detailed table with columns for compound name, molecular weight, averaged retention time (RT), Kovats index (or equivalent), chemical formula, CAS number, EI-MS match (total spectrum similarity percentage), simple/weight dot product, reverse dot product, compound coefficient of variance, and optionally, isotopic-standard validation. Moreover, the manuscript is based on the analysis of several metabolites. However, no normalization procedure for standardizing signal intensities was explicitly described. The frequency of running quality control samples, blank samples, and standards must be specified to assess any potential shifts in signal intensity or retention times that could impact sample comparisons.

Finally, the figures are of low quality, and Figure 2, the aroma profile spiderweb, is not presented in English.

[1] Summer et al. (2007). Metabolomics, 3(3), 211-221.

Comments on the Quality of English Language

The figures are of low quality (better resolution is required for publication), and Figure 2, the aroma profile spiderweb, is not presented in English.

Author Response

Comments 1： While the study is conceptually interesting, the authors base it entirely on the sensory evaluation conducted by a panel. It is not mentioned or demonstrated that this panel had professional training, as in the case of cacao, liquor, tea, etc., nor is the questionnaire used to qualify the taste provided.

Response 1：Three sensory training sessions were conducted. In the first one, panellists with rich experience in rice descriptive sensory analysis generated aroma descriptors for raw rice samples studied. For the second session, the panellists performed a sensory evaluation of the samples based on their own subjective perceptions and recorded some different aroma descriptors. In the last session, all sensory panellists were asked to participate in a discussion of sensory descriptors screening and deletion. Ultimately, teen aroma attributes (starchy, grainy, floral, Popcorn, sweet, grassy, woody, oatmeal, dairy, corn) were determined to be most effective in characterizing the rice and highlighting key differences among them. After the training period, the quantitative description analysis was performed using the screened sensory descriptors and the provided standard references. The strength of each descriptor ranged from 0 to 9, with 0 representing no sensation and 9 representing a strong sensation (0.5 is the minimum increment).

Comments 2：Furthermore, the manuscript currently lacks essential data and fails to adhere to the "Proposed Minimum Reporting Standards for Chemical Analysis" by the Chemical Analysis Working Group (CAWG) of the Metabolomics Standards Initiative (MSI). It should include a detailed table with columns for compound name, molecular weight, averaged retention time (RT), Kovats index (or equivalent), chemical formula, CAS number, EI-MS match (total spectrum similarity percentage), simple/weight dot product, reverse dot product, compound coefficient of variance, and optionally, isotopic-standard validation.

Response 2：A detailed table has been provided to illustrate the qualitative compounds, which includes information such as compound name, averaged retention time (RT), chemical formula, CAS number, category, odor threshold. This also refers to the latest research conducted by Zhang et al., which provides supplementary information on the relevant topics. [Zhang, X.; Dai, Z.; Fan, X.; Liu, M.; Ma, J.; Shang, W.; Zhou, Z. A study on volatile metabolites screening by HS-SPME-GC-MS and HS-GC-IMS for discrimination and characterization of white and yellowed rice. Cereal Chemistry.2020, 97(2),496-504. https://doi.org/10.1002/cche.10264]

Comments 3：Moreover, the manuscript is based on the analysis of several metabolites. However, no normalization procedure for standardizing signal intensities was explicitly described. The frequency of running quality control samples, blank samples, and standards must be specified to assess any potential shifts in signal intensity or retention times that could impact sample comparisons.

Response 3：The retention index (RI) was calculated using n-ketones C4-C9 (Sinopharm Chemical Reagent Beijing Co., Ltd., China) as external references by the Laboratory Analysis View (LAV) software in the GC-IMS device. The volatile compounds were tentative identified based on comparison of RI and the drift time with the NIST library and IMS database retrieval software obtained from G.A.S (Dortmund, Germany). Finally, the intensities of the volatile compounds were analyzed according to the peak volumes of the selected signal peaks by Gallery Plot analysis (v.1.0.7, G.A.S.) [32]. Visual and quantitative comparisons of volatility differences between samples were performed using the Gallery plot plug-in. Two-dimensional top views and three-dimensional fingerprints were constructed using the Reporter plug-in. In order to avoid deviations between measurements, the drift time of sample spectra was normalized relative to RIP drift time, which proceed automatically in the LAV software.

Comments 4：Finally, the figures are of low quality, and Figure 2, the aroma profile spiderweb, is not presented in English.

Response 4：The charts have been adjusted and optimized. Please refer to the latest submission version.

Reviewer 3 Report

Comments and Suggestions for Authors

Foods-3915485

The manuscript explores the molecular basis for the flavor differences of different rice aromas, analyzing the volatile organic compounds (VOCs) using HS-SPME-GC-MS and GC-IMS. Sensory evaluation categorized rice aromas into three types ( A, prominent sweet, popcorn; B, Cereal and starchy; and C, complex aroma), while instrumental analysis identified 291 VOCs. Comparing the aroma profiles of the three rice categories, identify the key aroma compounds that apparently contribute to each aroma type.

The topic may be relevant since rice is an essential food for China’s consumers. However, the special use made in China of this food made it difficult to extrapolate the results outside this country, which, in any case, represents a relevant proportion of the world population. The methodology is advanced and reliable, although some paragraphs could be clarified and concretized. Specifically, what was the sample analyzed (rice, processed rice…)? Regarding results, they describe exhaustively the different compounds found using the two analytical procedures, but also include comments that could be more suitable for the discussion section (e.g., those relating the compounds to the flavor imparted). The list of references is extensive and adequately supports the connections between compounds and the aroma perceived.  

L101. Brief information on the preparation procedures should be provided.

L116. The sentence should be clarified. Its current structure raises doubts about whether all samples are included.

L125-127. Should be reordered for clarification.

L130. What does “culturing” mean?

L132 Could the non-split injection cause any effect on the number of compounds detected and identified?

L142. Do you mean three analyses of the same sample? Were there replicates of processing? Was each sample processed only once?

 L144. Possibly, a brief information on ROAV could be necessary since this aspect could be relevant for readers.

L152. There is some confusion regarding the sample analyzed. What was the meaning of “prepared” (L101)?

L161-162. Are the results produced from one sample analyzed in triplicate?  Rice with the same rice classification could have different VOC profiles.

L173. Quantification method? Data are relative to the internal standard.

L186 Differences between spider plots and radar charts?

Figure 2. Legends are not in English.

Table 2. The table could also include the aroma class in which each compound was identified.

Figure 3. The placement of the identification letters below the figure may confuse. It should be better positioned at the top left or right.  

L234-235. The confidence ellipses are not shown in Fig. 3b

L244-246. It is supposed that ROAV correlates with the influence on the information in sensory analysis. It is possible that compounds in low proportions had a noticeable influence on the aroma perceived. Please comment on this possibility.

L249. Please, indicate which type of standardization was used.

Table 3. Would it be pertinent also to include columns indicating the aroma class where these compounds were found?

Fig. 4. Requires improvement. In its current state, the various graphs are not easily understandable; a similar issue arises with the placement of letters to differentiate between graphs in text.

L278-284. The explanation could be simplified by indicating compounds common to all groups, common to A and B, etc.

Figure 6. Legends are often challenging to understand and require improvement.

Table 4. Notice that in this table, aroma classes are included. It appears reasonable to include them in the previous tables as well. Do you agree?

Abbreviations. VIP, First principal component?

Author Response

Comments 1： L101. Brief information on the preparation procedures should be provided.

Response 1：This part only concerns sample collection.

Comments 2： L116. The sentence should be clarified. Its current structure raises doubts about whether all samples are included.

Response 2：Flavor metabolomic analysis was performed on all 11 different types of commercial raw rice. .

Comments 3： L125-127. Should be reordered for clarification.

Response 3：First, 0.816 μg/μL of 2‐methyl‐3‐heptanone standard solution was prepared using deionized water as the solvent. The rice samples (2000±40mg) were placed in 20-mL headspace bottles, and 1 μL of 2‐methyl‐3‐heptanone was added as an internal standard.

Comments 4： L130. What does “culturing” mean?

Response 4：Headspace injection conditions included incubation at 60°C, preheating for 15 min, adsorbing for 30 min, and desorption for 4 min.

Comments 5： L132 Could the non-split injection cause any effect on the number of compounds detected and identified?

Response 5：Non-split injection can enhance sensitivity and reduce the loss of samples in the chromatographic column. It is suitable for low-concentration samples.

Comments 6： L142. Do you mean three analyses of the same sample? Were there replicates of processing? Was each sample processed only once?

Response 6：Each experiment with the samples was conducted triplicately.

Comments 7：  L144. Possibly, a brief information on ROAV could be necessary since this aspect could be relevant for readers.

Response 7：The relative odor activity values (ROAV) were performed to evaluate the specific contribution of each compound to the overall aroma, which was well reported in previous studies[30]. ROAV was mainly used to quantitatively evaluate the contribution of each volatile substance to the overall flavor of the test sample and thus identify key aroma-active compounds.A high ROAV is also indicative of the great contribution of a component to the overall flavor of the sample. The compounds with ROAV ≥1 were the principal aroma compounds of the sample, whereas those with 0.1 ≤ ROAV < 1 had a important modifying effects on its overall flavor. ROAV was calculated as described previously.

Comments 8： L152. There is some confusion regarding the sample analyzed. What was the meaning of “prepared” (L101)?

Response 8：This part refers to the rice grinding process before the GC-IMS analysis. The earlier part mainly involves sample collection.

Comments 9： L161-162. Are the results produced from one sample analyzed in triplicate?  Rice with the same rice classification could have different VOC profiles.

Response 9：Each experiment with the same samples was conducted triplicately.

Comments 10： L173. Quantification method? Data are relative to the internal standard.

Response 10：Finally, the intensities of the volatile compounds were analyzed according to the peak volumes of the selected signal peaks by Gallery Plot analysis (v.1.0.7, G.A.S.) . Visual and quantitative comparisons of volatility differences between samples were performed using the Gallery plot plug-in.

Comments 11： L186 Differences between spider plots and radar charts?

Response 11：Spider plots and radar charts have similar meanings. Keep the "spider plots".

Comments 12： Figure 2. Legends are not in English.

Response 12：The charts have been adjusted and optimized. Please refer to the latest submission version.

Comments 13： Table 2. The table could also include the aroma class in which each compound was identified.

Response 13：The table have been adjusted and optimized. Please refer to the latest submission version.

Comments 14： Figure 3. The placement of the identification letters below the figure may confuse. It should be better positioned at the top left or right.  

Response 14：The figure have been adjusted and optimized. Please refer to the latest submission version. The letters in the bottom of the picture were taken as references from the provided template.

Comments 15： L234-235. The confidence ellipses are not shown in Fig. 3b

Response 15：The figure have been adjusted and optimized. Please refer to the latest submission version.

Comments 16： L244-246. It is supposed that ROAV correlates with the influence on the information in sensory analysis. It is possible that compounds in low proportions had a noticeable influence on the aroma perceived. Please comment on this possibility.

Response 16：As shown in the research conducted by Gemert[Gemert, L.J.V. Compilations of Odour Threshold Values in Air, Water and Other Media.], ROAV was mainly used to quantitatively evaluate the contribution of each volatile substance to the overall flavor of the test sample and thus identify key aroma-active compounds.A high ROAV is also indicative of the great contribution of a component to the overall flavor of the sample. The compounds with ROAV ≥1 were the principal aroma compounds of the sample, whereas those with 0.1 ≤ ROAV < 1 had a important modifying effects on its overall flavor. 

Comments 17： L249. Please, indicate which type of standardization was used.

Response 17： In order to avoid deviations between measurements, standardization proceed automatically in the software.

Comments 18： Table 3. Would it be pertinent also to include columns indicating the aroma class where these compounds were found?

Response 18：The table have been adjusted and optimized. Please refer to the latest submission version.

Comments 19： Fig. 4. Requires improvement. In its current state, the various graphs are not easily understandable; a similar issue arises with the placement of letters to differentiate between graphs in text.

Response 19：The figure have been adjusted and optimized. Please refer to the latest submission version.

Comments 20： L278-284. The explanation could be simplified by indicating compounds common to all groups, common to A and B, etc.

Response 20：The explanation have been adjusted and optimized. Please refer to the latest submission version.

Comments 21： Figure 6. Legends are often challenging to understand and require improvement.

Response 21：The figure have been adjusted and optimized. Please refer to the latest submission version.

Comments 22： Table 4. Notice that in this table, aroma classes are included. It appears reasonable to include them in the previous tables as well. Do you agree?

Response 22：The table have been adjusted and optimized. Please refer to the latest submission version.

Comments 23： Abbreviations. VIP, First principal component?

Response 23：VIP, Variable Importance in the Projection.