Method for distinguishing regional flavor of pickled mustard based on combination of characteristic volatile compounds

Abstract

This invention relates to the technical field of food flavor analysis and quality evaluation, and particularly to a method for regional flavor discrimination of convenient pickled mustard tuber based on a combination of characteristic volatile compounds. This method uses a reference sample and a sample to be tested to extract a feature vector composed of exogenous spice markers and endogenous markers of the pickled mustard tuber. Exogenous markers include anethole, linalool, and at least one selected from 4-allyl-2-methoxyphenol, α-terpineol, and 4-terpene alcohol. Endogenous markers include nonanal and / or phenylethyl isothiocyanate. After logarithmic transformation of the content, an orthogonal partial least squares discriminant analysis model is established. The feature vector of the sample to be tested is input, and the output is Fuling type, Ningbo type, Meishan type, or unknown / mixed type. This invention uses a dual-source compound combination of "exogenous flavoring characteristics - endogenous raw material characteristics" instead of a single electronic nose signal or all volatile compounds, which can reduce the interference of complex flavoring formulas and high-dimensional redundant variables on the discrimination and improve the chemical interpretability of the discrimination results.

Description

Technical Field

[0001] This invention relates to the technical field of food flavor analysis and quality evaluation, and in particular to a method for determining the regional flavor of convenient pickled mustard tubers based on combinations of characteristic volatile compounds. Background Technology

[0002] Convenient pickled mustard tubers are typically made from pickled mustard tuber embryos through processes such as cutting, desalting, seasoning, packaging, and sterilization. Differences in raw material varieties, fermentation environment, salting processes, and seasoning formulas collectively influence the volatile flavor composition of the finished product, resulting in relatively stable product styles across different production regions. Objectively identifying regional flavor types helps companies establish reference flavor libraries, control batch fluctuations, and evaluate product style deviations.

[0003] Current methods for evaluating the flavor of convenient pickled mustard tubers largely rely on human sensory evaluation, conventional physicochemical indicators, electronic noses, or full-spectrum gas chromatography-mass spectrometry (GC-MS). Human sensory evaluation is easily affected by the evaluator's condition and training level; electronic noses can quickly acquire overall odor signals, but the sensor response lacks clear directionality at the compound level, and samples from different regions may overlap on some principal component axes; while full-spectrum GC-MS can obtain a large number of volatile compounds, directly using all variables to build a discriminant model can easily introduce redundant variables and reduce the model's interpretability and ease of subsequent generalization.

[0004] Convenient pickled mustard tubers differ from unseasoned primary agricultural products. The aroma of the finished product contains both exogenous volatile components from spices such as star anise, fennel, and Sichuan peppercorns, and endogenous components related to the raw materials and fermentation process, such as nonanal, nitriles, and isothiocyanates. Using only spice-derived compounds or only endogenous compounds as discriminant variables is easily affected by variations in company formulations or differences in fermentation batches. Therefore, it is necessary to utilize a combination of dual-source markers—both exogenous flavoring characteristics and endogenous raw material characteristics—to obtain regional flavor type discrimination results with a chemical explanation basis.

[0005] Existing methods for identifying food origin or variety often employ volatile fingerprinting and principal component analysis, partial least squares discriminant analysis, or orthogonal partial least squares discriminant analysis. However, there is a lack of technical solutions that, for the complex flavor matrix of convenient pickled mustard tubers, define exogenous spice markers such as anethole and linalool together with endogenous characteristic substances such as nonanal and phenylethyl isothiocyanate as a characteristic combination, and then determine the regional flavor type accordingly. Summary of the Invention

[0006] The purpose of this invention is to provide a method for determining the regional flavor of convenient pickled mustard tubers based on combinations of characteristic volatile compounds, in order to solve the problems mentioned in the background art, such as the complexity of the flavoring components of convenient pickled mustard tubers, the insufficient interpretability of electronic nose signals, and the excessive number of full-spectrum volatile variables.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a convenient method for identifying regional flavors of pickled mustard tuber based on characteristic volatile compound combinations, comprising the following steps:

[0008] S1. Obtain reference samples of convenient pickled mustard tubers with known regional flavor types and samples of convenient pickled mustard tubers to be tested;

[0009] S2. The reference sample and the sample to be tested are respectively pulverized, subjected to headspace solid-phase microextraction and gas chromatography-mass spectrometry to obtain the qualitative results and content data of volatile compounds in each sample.

[0010] S3. Extract a characteristic volatile compound combination from the obtained volatile compounds to form a sample feature vector. The characteristic volatile compound combination simultaneously includes exogenous spice markers and endogenous markers of pickled mustard tuber. The exogenous spice markers include anethole, linalool, and at least one selected from 4-allyl-2-methoxyphenol, α-terpineol, and 4-terpene alcohol. The endogenous markers of pickled mustard tuber include at least one selected from nonanal and phenethyl isothiocyanate.

[0011] S4. Perform logarithmic transformation on the content data in the feature vector of the reference sample, establish an initial OPLS-DA model with regional flavor type as the classification variable, calculate the VIP value of each feature compound, screen feature compounds with VIP>1, and establish the final discrimination model based on the screened variables.

[0012] S5. Obtain the feature vector of the sample to be tested according to steps S2 and S3 and the data conversion rules consistent with the reference sample, and input it into the established model. Input it into the established orthogonal partial least squares discriminant analysis model and output the regional flavor type of the sample to be tested.

[0013] Furthermore, the orthogonal partial least squares discriminant analysis model can adopt a multi-class model, or a one-to-one or one-to-the-others multiple binary classification models. The model output types include at least the Chongqing Fuling type, the Zhejiang Ningbo type, and the Sichuan Meishan type. For samples exceeding the Hotelling T² range, the DModX threshold, or the 95% confidence range, the output type can be unknown or mixed.

[0014] Preferably, in the characteristic volatile compound combination, the exogenous spice markers further include at least one of artemisinin, trans-anetinoside, α-pinene, β-caryophyllene, and D-limonene, and the endogenous markers of pickled mustard tuber further include at least one of a nitrile compound (phenylpropionitrile) and a sulfur-containing compound (allyl isothiocyanate).

[0015] Preferably, the regional flavor types include Chongqing Fuling type, Zhejiang Ningbo type and Sichuan Meishan type. When the sample to be tested exceeds the applicable domain of the model established based on the reference sample, the output is unknown or mixed type. The Chongqing Fuling type, Zhejiang Ningbo type and Sichuan Meishan type are flavor chemical classification labels formed under unified analytical conditions and a predetermined reference sample library, and are not directly equivalent to the conclusion of the authenticity of the legal geographical origin, raw material origin or geographical indication.

[0016] Preferably, each regional flavor type in the reference sample set contains at least six independent commodity samples with clearly defined sources. The main ingredients of each reference sample include salt, vegetable oil, chili pepper, spices, and monosodium glutamate, and the production dates of each reference sample are within the same preset time window.

[0017] Preferably, in step S2,

[0018] For HS-SPME pretreatment, weigh 3.00 g of the pulverized sample and place it in a 20 mL headspace vial. Add 10 μL of octanol internal standard solution with a mass concentration of 8.19 μg / mL, equilibrate at 60 °C for 20 min, and extract with 50 / 30 μm DVB / CAR / PDMS extraction fiber at 60 °C for 40 min.

[0019] GC-MS chromatographic and mass spectrometric conditions were applied, followed by resolution at 250℃ for 5 min at the injection port of the gas chromatograph-mass spectrometer. The gas chromatographic column was a 30m×0.25mm×0.25μm 5% phenyl-95% dimethyl polysiloxane capillary column, with helium as the carrier gas and a flow rate of 1.11 mL / min. The temperature program was as follows: 50℃ for 4 min, increased to 110℃ at 3℃ / min and held for 2 min, increased to 150℃ at 3℃ / min and held for 6 min, and finally increased to 230℃ at 5℃ / min and held for 2 min. The mass spectrometry used an electron impact ion source with an electron energy of 70 eV and an ion source temperature of 230℃. The scan range was 50-550 m / z.

[0020] Preferably, the volatile compounds are qualitatively identified by mass spectrometry library retrieval and retention index, and the content M is calculated using an internal standard semi-quantitative method. The calculation formula is M=[ρ×V×Aᵢ / (A s [×m)]×10³; where M is the volatile compound content, ρ is the internal standard mass concentration, V is the internal standard volume, Aᵢ is the peak area of ​​the target volatile compound, and A s denoted as , where m is the peak area of ​​the internal standard and m is the sample mass.

[0021] Preferably, in step S4, the independent variable fit index R²X, dependent variable fit index R²Y, and prediction index Q² of the established orthogonal partial least squares discriminant analysis model are all greater than 0.5, and the overfitting risk of the model is evaluated by at least 200 permutation tests, wherein the intersection of the Q² regression line and the vertical axis of the permutation test is less than 0.

[0022] Preferably, in step S3, anethole is used as the exogenous core differential marker, and linalool, 4-allyl-2-methoxyphenol, α-terpineol, 4-terpene alcohol, artemisinin and trans-anenethole are used to supplement the exogenous spice profile. Nonanal, phenylethyl isothiocyanate, nitrile compounds (phenylpropionitrile) and sulfur-containing compounds (allyl isothiocyanate) are used to jointly reflect the endogenous green aroma and spiciness characteristics of pickled mustard tuber. The exogenous spice markers and the endogenous markers of pickled mustard tuber are jointly input into the orthogonal partial least squares discriminant analysis model.

[0023] Preferably, before step S2, an electronic nose pre-screening step is included: 3.00g of the pulverized sample is weighed and placed in a 20mL headspace vial, and heated in a 60℃ water bath for 15min. The sample is collected for 90s and washed for 120s at a gas flow rate of 1L / min. The response values ​​of 10 sensors, namely W1C, W5S, W3C, W6S, W5C, W1S, W1W, W2S, W2W and W3S, are obtained and principal component analysis is performed to determine whether the sample to be tested is within the overall odor distribution range of the reference sample.

[0024] Compared with the prior art, the beneficial effects of the present invention are:

[0025] 1. This invention uses a combination of characteristic volatile compounds composed of exogenous spice markers and endogenous pickled mustard tuber markers as the discriminant variable, rather than a single electronic nose signal or a spectrum of all volatile compounds. This not only reflects the style of the exogenous seasoning formula and the fermentation characteristics of the endogenous raw materials, but also has clear compound-level orientation and sensory significance. This effectively reduces the interference of complex seasoning matrices on the discrimination results and reduces the risk of misjudgment caused by relying on a single source.

[0026] 2. This invention standardizes the sample pretreatment steps, including standardized sample volume, headspace equilibrium conditions, extraction fibers, extraction time, and gas chromatography-mass spectrometry detection conditions. It also uses an internal standard semi-quantitative method to calculate the content, which helps to reduce systematic errors caused by operational differences and improves the reproducibility and comparability of detection results.

[0027] 3. This invention effectively weakens the singular dominance of high-concentration spice components on the model by performing logarithmic transformation on the content of volatile compounds and screening key variables by combining variable importance projection values ​​(VIP>1). At the same time, it eliminates redundant variables with low contribution, reduces the risk of model overfitting, and helps to reduce variable redundancy and model overfitting risk, thereby improving the chemical interpretability of the model.

[0028] 4. This invention allows electronic noses to be used as a rapid pre-screening method for chemical confirmation using characteristic combinations of volatile compounds. It utilizes the high efficiency of overall odor analysis of electronic noses while overcoming their lack of specific compound targeting, thus balancing detection efficiency and result interpretability. It is suitable for rapid screening in industrial settings and confirmatory analysis in laboratories.

[0029] 5. The regional flavor types output by this invention (including Chongqing Fuling type, Zhejiang Ningbo type, Sichuan Meishan type, etc.) are product aroma chemical style classification labels formed under specific reference sample libraries and unified analytical conditions. For samples that exceed the model confidence range, the output is unknown or mixed. It is clearly not directly equivalent to the conclusion of the legal geographical origin, raw material origin or authenticity of geographical indication. The technical boundaries are reasonably defined, the technical application boundaries of this method are clarified, and the flavor classification results are avoided from being directly interpreted as the conclusion of the geographical origin or authenticity of geographical indication. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the method for conveniently identifying regional flavor types of pickled mustard tuber according to the present invention;

[0031] Figure 2 This is a schematic diagram of principal component analysis of the electronic nose response value of the reference sample in an embodiment of the present invention, where A is the score plot and B is the loading plot;

[0032] Figure 3 This is a schematic diagram showing the PCA scores of volatile compounds in convenient pickled mustard tubers in different regions according to an embodiment of the present invention;

[0033] Figure 4 These are two sets of OPLS-DA score charts in an embodiment of the present invention, where A represents Chongqing Fuling vs. Zhejiang Ningbo, and B represents Sichuan Meishan vs. Zhejiang Ningbo;

[0034] Figure 5 These are two sets of OPLS-DA permutation test plots in an embodiment of the present invention, where C represents Chongqing Fuling vs. Zhejiang Ningbo, and D represents Sichuan Meishan vs. Zhejiang Ningbo;

[0035] Figure 6 These are two sets of OPLS-DA VIP diagrams in an embodiment of the present invention, where E represents Chongqing Fuling vs. Zhejiang Ningbo, and F represents Sichuan Meishan vs. Zhejiang Ningbo. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Unless otherwise specified, each experiment shall be performed at least three times, and the results shall be expressed as mean ± standard deviation.

[0037] Example 1: Establishment of a Reference Sample Set

[0038] Thirty-five commercially available ready-to-eat pickled mustard tubers were selected as reference samples of known types. These included 20 samples from Fuling District, Chongqing (numbered Y1-Y20); 6 samples from Ningbo City, Zhejiang Province (numbered Z1-Z6); and 9 samples from Meishan City, Sichuan Province (numbered C1-C9). All samples were sourced from large supermarket chains or official online channels of the brands, were in intact packaging, and had a production date of May 2025.

[0039] To minimize the impact of differences in enterprise formulations, reference samples were preferentially selected from products with similar ingredient lists, whose main ingredients all included salt, vegetable oil, chili peppers, spices, and monosodium glutamate. Samples were stored at 25°C until analysis. The distribution of reference samples is shown in Table 1.

[0040] Table 1 Reference Sample Information

[0041]

[0042] Example 2: Electronic nose for overall odor pre-screening

[0043] Accurately weigh 3.00g of the pulverized ready-to-use pickled mustard tuber sample and place it in a 20mL headspace vial. Incubate in a 60℃ water bath for 15min. The electronic nose cleaning time is 120s, the data acquisition time is 90s, and the gas flow rate is 1L / min. The electronic nose sensor and main response objects are shown in Table 2.

[0044] Table 2 Electronic nose sensors and main response objects

[0045]

[0046] Principal component analysis (PC Component Analysis) was performed using the response values ​​of 10 sensors. PC1 and PC2 contributed 79.40% and 14.60%, respectively, with a cumulative contribution of 94.00%. Samples from the Chongqing Fuling type were mainly located near sensors W1C, W5S, and W3C; samples from the Zhejiang Ningbo type and the Sichuan Meishan type were mainly located near sensors W2S, W2W, and W3S. The electronic nose results indicate regional differences in overall sample odor, but different types may still overlap in some principal component directions. Therefore, the electronic nose is preferred as a pre-screening criterion rather than the sole criterion.

[0047] Example 3: HS-SPME-GC-MS Detection of Volatile Compounds

[0048] Take 3.00 g of pulverized ready-to-use pickled mustard tuber sample and place it in a 20 mL headspace solid-phase microextraction (HSP) vial. Add 10 μL of 8.19 μg / mL 2-octanol internal standard solution, seal the vial, and equilibrate at 60 °C for 20 min. Insert a 50 / 30 μm DVB / CAR / PDMS extraction fiber into the headspace vial and extract at 60 °C for 40 min. Immediately after extraction, insert the fiber into the GC-MS inlet and elute at 250 °C for 5 min.

[0049] Gas chromatography conditions: SH-I-5SilMS or equivalent stationary phase capillary column, 30m × 0.25mm × 0.25μm; splitless injection; high-purity helium as carrier gas, flow rate 1.11mL / min, column pressure 107.9kPa. Temperature program: 50℃ for 4 min, increase to 110℃ at 3℃ / min and hold for 2 min, increase to 150℃ at 3℃ / min and hold for 6 min, increase to 230℃ at 5℃ / min and hold for 2 min.

[0050] Mass spectrometry conditions: electron impact ionization source, ion source temperature 230℃, electron energy 70 eV, scan range 50-550 m / z, detector voltage 0.1 kV. A mixed standard of C7-C40 n-alkanes was detected under the same conditions, and retention indices were calculated. Target compounds were identified using NIST mass spectrometry library matching and retention indices.

[0051] Volatile compounds were semi-quantitatively analyzed using the internal standard method, calculated using the following formula: M = [ρ × V × Ai / (A s [×m)]×10³. Where M is the content of the target compound, in μg / kg; ρ is the mass concentration of the internal standard, in μg / mL; V is the volume of the internal standard, in mL; Ai is the peak area of ​​the target compound; A s denoted as the internal standard peak area; m represents the sample mass in grams.

[0052] Example 4: Volatile Flavor Characteristics of Reference Sample

[0053] A total of 146 volatile flavor compounds were identified in 35 reference samples, including 34 esters, 18 aldehydes, 10 ketones, 20 alcohols, 2 nitriles, 2 acids, 6 sulfur-containing compounds, 39 terpenes, 5 phenols, 5 ethers, and 5 other compounds.

[0054] Ethers were the main volatile components in the finished ready-to-eat pickled mustard tubers from all three regions. The average ether content of samples from Chongqing Fuling, Zhejiang Ningbo, and Sichuan Meishan was 692.19, 1070.57, and 536.94 μg / kg, respectively. Representative compounds included anethole, trans-anenethole, and artemisinin. Linalool was the main alcohol; terpenes mainly included α-pinene, β-caryophyllene, and D-limonene. These compounds primarily reflect the exogenous spice flavoring characteristics developed during the finished product's seasoning process.

[0055] The average total volatile compounds in samples from Chongqing Fuling, Zhejiang Ningbo, and Sichuan Meishan were 986.35, 1384.60, and 781.82 μg / kg, respectively. The total volatile compounds in the Zhejiang Ningbo samples ranged from 1221.08 to 1465.78 μg / kg, showing relatively small fluctuations.

[0056] Regarding the endogenous characteristics of pickled mustard tuber, the average aldehyde content of the Meishan type samples from Sichuan was 26.03 μg / kg, with nonanal being the representative substance; the average sulfur compound content of the Fuling type samples from Chongqing was 21.92 μg / kg, which was 3.43 times that of the Ningbo type from Zhejiang and 1.96 times that of the Meishan type from Sichuan, respectively, and also showed a high nitrile content. This demonstrates that the combined characterization of exogenous spice components and endogenous aromatic and pungent components can reflect the flavor types of different regions.

[0057] Example 5: Key aroma active substances and dual-source characteristic combinations

[0058] Odor activity value (OAV) was calculated as the ratio of the target compound's content to its odor threshold. When the OAV was greater than 1, the compound was considered a key aroma active substance that directly contributed to the overall aroma of the sample. A total of 21 compounds with an OAV greater than 1 were screened from samples of three different regions.

[0059] Linalool, artemisia argyi, trans-anetinoside, and anethole generally have an OAV greater than 1 in ready-to-eat pickled mustard tubers. The average OAV of linalool in samples from Chongqing Fuling, Zhejiang Ningbo, and Sichuan Meishan were 11.34, 18.67, and 17.11, respectively. In the Sichuan Meishan samples, except for C4, nonanal had an OAV of 1.79-18.54, exhibiting a green aroma; nearly half of the Chongqing Fuling samples had an OAV greater than 1 for phenethyl isothiocyanate, exhibiting a pungent characteristic.

[0060] Based on the above results, this embodiment forms the dual-source feature combination shown in Table 3. This combination includes both exogenous markers reflecting the flavoring style of spices and endogenous markers reflecting the raw materials and fermentation retention characteristics of pickled mustard tuber.

[0061] Table 3 Combinations of exogenous and endogenous characteristic flavor compounds

[0062]

[0063] Example 6: PCA and OPLS-DA Model Establishment and Biomarker Screening

[0064] Logarithmically transform the volatile compound content of each sample to reduce the impact of different concentration orders on the model. Undetected or zero values ​​can be handled using the ln(x+1) form, or missing value substitution methods commonly used in the field can be used according to the detection limit, but the reference sample and the sample to be tested must be processed using the same rules.

[0065] Principal component analysis was first performed on 146 volatile compounds. PC1 and PC2 contributed 45.70% and 16.00% respectively, explaining a total of 61.70% of the information. Samples from Ningbo, Zhejiang clustered relatively tightly; samples from Fuling, Chongqing had a wider distribution; samples from Meishan, Sichuan were mainly distributed near the negative PC1 and PC2 indices, showing some distinction from the Ningbo, Zhejiang samples.

[0066] Further, orthogonal partial least squares discriminant analysis models were established. Based on existing data, binary classification models could be established for the Chongqing Fuling type and the Zhejiang Ningbo type, as well as the Sichuan Meishan type and the Zhejiang Ningbo type. One model had R²X of 0.790, R²Y of 0.960, and Q² of 0.768, all greater than 0.5. After 200 permutation tests, the intersection of the Q² regression line and the vertical axis was less than 0.

[0067] Using a VIP greater than 1 to screen key variables, eight key differentially expressed volatile compounds were identified between the Chongqing Fuling type and the Zhejiang Ningbo type, and six key differentially expressed volatile compounds were identified between the Sichuan Meishan type and the Zhejiang Ningbo type. Anethole had VIP values ​​of 7.78 and 7.09 in the two comparisons, respectively; 4-allyl-2-methoxyphenol, linalool, α-terpineol, and 4-terpene alcohol also showed high discriminative contributions.

[0068] In formal applications, a three-class OPLS-DA model can be established based on the category labels of the reference samples, or three one-to-one models can be established: Fuling-Ningbo, Meishan-Ningbo, and Fuling-Meishan, with the results output using majority voting. For samples that do not fall within the 95% confidence range of any reference type, an unknown or mixed type is output to avoid forcibly classifying samples outside the model.

[0069] Example 7: Sample discrimination process

[0070] Step 1: Obtain the ready-to-eat pickled mustard tuber to be tested, and confirm that it is a ready-to-eat pickled mustard tuber product formed by desalting, seasoning and packaging; record the ingredient list, production batch and sample status.

[0071] Step 2: Perform HS-SPME-GC-MS detection according to the unified method in Example 3 to obtain the contents of anethole, linalool, 4-allyl-2-methoxyphenol, α-terpineol, 4-terpene alcohol, phenylpropionitrile, allyl isothiocyanate, nonanal and phenylethyl isothiocyanate.

[0072] Step 3: Construct the feature vector of the test sample using the same logarithmic transformation and standardized parameters as the reference sample. The standardized parameters can only be calculated using the reference sample; the mean and standard deviation must not be recalculated using the test sample.

[0073] Step 4: Input the feature vector of the sample to be tested into the established OPLS-DA model, and read the predicted category, prediction score, Hotelling T² and DModX results.

[0074] Step 5: When the predicted category is clear and the sample is within the applicable range of the model, output the corresponding regional flavor type; when multiple model results conflict or exceed the applicable range of the model, output unknown or mixed type, and prompt for re-determination after expanding the reference sample library.

[0075] Comparative Example 1: Discrimination using only electronic nose PCA

[0076] When PCA was performed using only the response values ​​of the 10 sensors from the electronic nose, although the cumulative contribution rate of the first two principal components reached 94.00%, there was overlap in the PC1 direction among samples from different regions, and the electronic nose could not provide an explanation for the differences at the specific compound level. Therefore, while the electronic nose can be used for rapid pre-screening, it should not be the sole confirmatory step for regional flavor type identification in this invention.

[0077] Comparative Example 2: Discrimination using only exogenous spice compounds

[0078] If only exogenous spice compounds such as anethole, linalool, and terpenes are selected, the model mainly reflects the differences in enterprise seasoning formulations and is difficult to distinguish the degree of retention of raw materials and fermentation characteristics. This invention further adds nonanal, phenylethyl isothiocyanate, allyl isothiocyanate, and phenylpropionitrile, so that the model can simultaneously reflect the endogenous characteristics of green aroma and spiciness. This comparative example is only used for technical logic comparison and does not limit the specific accuracy difference between the two models.

[0079] Industrial applicability

[0080] This invention can be implemented using headspace solid-phase microextraction equipment, gas chromatography-mass spectrometry, and chemometrics software commonly found in food flavor laboratories. Once the reference sample library is established, it can be used to confirm the flavor type of new batches of samples for enterprises, identify product formulation deviations, compare product styles from different regions, and standardize quality, thus meeting the requirements for industrial application.

[0081] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A convenient method for identifying regional flavors of pickled mustard tuber based on characteristic volatile compound combinations, characterized in that, Includes the following steps: S1. Obtain reference samples of convenient pickled mustard tubers with known regional flavor types and samples of convenient pickled mustard tubers to be tested; S2. The reference sample and the sample to be tested are respectively pulverized, subjected to headspace solid-phase microextraction and gas chromatography-mass spectrometry to obtain the qualitative results and content data of volatile compounds in each sample. S3. Extract a characteristic volatile compound combination from the obtained volatile compounds to form a sample feature vector. The characteristic volatile compound combination simultaneously includes exogenous spice markers and endogenous markers of pickled mustard tuber. The exogenous spice markers include anethole, linalool, and at least one selected from 4-allyl-2-methoxyphenol, α-terpineol, and 4-terpene alcohol. The endogenous markers of pickled mustard tuber include at least one selected from nonanal and phenethyl isothiocyanate. S4. Perform logarithmic transformation on the content data in the feature vector of the reference sample, establish an initial OPLS-DA model with regional flavor type as the classification variable, calculate the VIP value of each feature compound, screen feature compounds with VIP>1, and establish a final discrimination model based on the screened variables. S5. Obtain the feature vector of the sample to be tested according to steps S2 and S3 and the data conversion rules consistent with the reference sample, and input it into the established model. Input it into the established orthogonal partial least squares discriminant analysis model and output the regional flavor type of the sample to be tested.

2. The convenient method for identifying regional flavors of pickled mustard tuber based on characteristic volatile compound combinations according to claim 1, characterized in that, The characteristic volatile compound combination includes, among other things, exogenous spice markers, at least one of artemisinin, trans-anetinoside, α-pinene, β-caryophyllene, and D-limonene, and endogenous markers of pickled mustard tuber include at least one of nitrile compounds (phenylpropionitrile) and sulfur-containing compounds (allyl isothiocyanate).

3. The convenient method for identifying regional flavors of pickled mustard tuber based on characteristic volatile compound combinations according to claim 1, characterized in that, The regional flavor types include Chongqing Fuling type, Zhejiang Ningbo type, and Sichuan Meishan type. When the sample to be tested exceeds the applicable domain of the model established based on the reference sample, the output is unknown or mixed type. The Chongqing Fuling type, Zhejiang Ningbo type, and Sichuan Meishan type are flavor chemical classification labels formed under unified analytical conditions and a predetermined reference sample library, and are not directly equivalent to the conclusion of the authenticity of the legal geographical origin, raw material origin, or geographical indication.

4. The convenient method for identifying regional flavors of pickled mustard tuber based on characteristic volatile compound combinations according to claim 1, characterized in that, The reference sample set contains at least six independent product samples with clearly identified sources for each regional flavor type. The main ingredients of each reference sample include salt, vegetable oil, chili pepper, spices, and monosodium glutamate, and the production dates of each reference sample are within the same preset time window.

5. The convenient method for identifying regional flavors of pickled mustard tuber based on characteristic volatile compound combinations according to claim 1, characterized in that, In step S2, For HS-SPME pretreatment, weigh 3.00 g of the pulverized sample and place it in a 20 mL headspace vial. Add 10 μL of octanol internal standard solution with a mass concentration of 8.19 μg / mL, equilibrate at 60 °C for 20 min, and extract with 50 / 30 μm DVB / CAR / PDMS extraction fiber at 60 °C for 40 min. GC-MS chromatographic and mass spectrometric conditions were applied, followed by resolution at 250℃ for 5 min at the injection port of the gas chromatograph-mass spectrometer. The gas chromatographic column was a 30m×0.25mm×0.25μm 5% phenyl-95% dimethyl polysiloxane capillary column. The carrier gas was helium, and the flow rate was 1.11 mL / min. The temperature program was as follows: 50℃ for 4 min, increased to 110℃ at 3℃ / min and held for 2 min, increased to 150℃ at 3℃ / min and held for 6 min, and finally increased to 230℃ at 5℃ / min and held for 2 min. The mass spectrometry used an electron impact ion source with an electron energy of 70 eV and an ion source temperature of 230℃. The scan range was 50-550 m / z.

6. The convenient method for identifying regional flavors of pickled mustard tuber based on characteristic volatile compound combinations according to claim 1, characterized in that, The volatile compounds were qualitatively identified by mass spectrometry library retrieval and retention index, and their content M was calculated using an internal standard semi-quantitative method. The calculation formula is M=[ρ×V×Aᵢ / (A s [×m)]×10³, where M is the volatile compound content, ρ is the internal standard mass concentration, V is the internal standard volume, Aᵢ is the peak area of ​​the target volatile compound, and A s denoted as , where m is the peak area of ​​the internal standard and m is the sample mass.

7. The convenient method for identifying regional flavors of pickled mustard tuber based on characteristic volatile compound combinations according to claim 1, characterized in that, In step S4, the independent variable fit index R²X, dependent variable fit index R²Y, and prediction index Q² of the established orthogonal partial least squares discriminant analysis model are all greater than 0.5, and the overfitting risk of the model is evaluated by at least 200 permutation tests, wherein the intersection of the Q² regression line and the vertical axis of the permutation test is less than 0.

8. The convenient method for identifying regional flavors of pickled mustard tuber based on characteristic volatile compound combinations according to claim 1, characterized in that, In step S3, anethole is used as the exogenous core differential marker. At least one of linalool, 4-allyl-2-methoxyphenol, α-terpineol, 4-terpene alcohol, artemisia argyi, and trans-anenethole is used to supplement the exogenous spice profile. At least one of nonanal, phenylethyl isothiocyanate, nitrile compounds (phenylpropionitrile), and sulfur-containing compounds (allyl isothiocyanate) is used to jointly reflect the endogenous green aroma and spiciness characteristics of pickled mustard tuber. The exogenous spice markers and the endogenous markers of pickled mustard tuber are jointly input into the orthogonal partial least squares discriminant analysis model.

9. The convenient method for identifying regional flavors of pickled mustard tuber based on characteristic volatile compound combinations according to claim 1, characterized in that, Before step S2, an electronic nose pre-screening step is also included: 3.00g of the pulverized sample is weighed and placed in a 20mL headspace vial, and heated in a 60℃ water bath for 15min. The sample is collected for 90s and washed for 120s at a gas flow rate of 1L / min. The response values ​​of 10 sensors, namely W1C, W5S, W3C, W6S, W5C, W1S, W1W, W2S, W2W and W3S, are obtained and principal component analysis is performed to determine whether the sample to be tested is within the overall odor distribution range of the reference sample.

10. The convenient method for identifying regional flavors of pickled mustard tuber based on characteristic volatile compound combinations according to any one of claims 1 to 9, characterized in that, The method is used to facilitate the evaluation of regional flavor styles of pickled mustard tubers, product grading, batch consistency monitoring, identification of seasoning formula deviations, or updating of reference sample libraries.

Images