Method for identifying invagine honey based on metabonomics technology

Abstract

The invention discloses a metabonomics technology-based method for identifying invar terrace honey, which comprises the following steps of: respectively detecting the invar terrace honey and known group honey by using UPLC-Q-TOF-MS (Ultra Performance Liquid Chromatography-Quadrature Time of Flight Mass Spectrometry) to obtain data of the invar terrace honey and data of the known group honey, preprocessing the data of the invar terrace honey and the data of the known group honey, and determining the metabonomics technology-based method for identifying the invar terrace honey. Respectively obtaining compound components of the invagine honey and the known group of honey; carrying out principal component analysis and orthogonal partial least square discriminant analysis on the compound components of the invagine honey and the known group of honey, and screening differential metabolites of the invagine honey and the known group of honey through a V + S graph; establishing a PLS discrimination model according to the differential metabolites; determining the differential metabolites of the to-be-detected honey, substituting the differential metabolites into the PLS discrimination model to obtain the predicted value of the to-be-detected honey, and then determining whether the to-be-detected honey is the invar terrace honey or not according to the preset predicted value range; according to the method, the characteristic marker of the invagine honey is determined, and the discrimination model is established, so that the authenticity of the invagine honey is effectively identified, and the accuracy is high.

Description

Technical Field

[0001] The invention relates to the technical field of food adulteration, and in particular to a method for identifying Yinjiaping honey based on metabolomics technology. Background Art

[0002] Honey, as a natural sweetener that can be used as both medicine and food, has high nutritional and medicinal value. In particular, high-quality specialty honey is deeply recognized and favored by consumers.

[0003] "Yinjiaping Honey" is a special honey that gathers the essence of wild plants in the mountains, especially a large number of Chinese herbal medicine stamens, and is rich in nutritional active ingredients. The production of "Yinjiaping Honey" strictly follows the ancient method, uses the traditional beekeeping tools of "Three Gorges Bucket", adheres to the custom of harvesting honey around the White Dew Festival of the lunar calendar, and only takes honey once a year. It insists on the process of harvesting honey by cutting the honeycomb and squeezing it. It has unique quality and flavor. As a high-quality special honey in my country, it has great market value.

[0004] In reality, it is common to add fructose syrup, starch syrup, and rice syrup to real honey to fake honey. Since the ratio of fructose to glucose in these syrups is very similar to that in honey, the various test indicators after addition are in full compliance with national standards, which makes detection very difficult. In addition, the use of low-priced honey of other varieties to impersonate high-priced honey is more difficult to identify and distinguish than simply adding syrup. The methods commonly used for identifying honey adulteration include stable carbon isotope ratio analysis (SCIRA), thin layer chromatography (TIC), high performance anion exchange chromatography with pulsed current detector (HPAEC-PAD), gas chromatography-mass spectrometry (GC-MS), high performance liquid chromatography (HPLC), high performance liquid phase isotope mass spectrometry (HPLC-IRMS), nuclear magnetic resonance (NMR), and near infrared spectroscopy (NIRS). The existing methods have many disadvantages in determining honey adulteration or the phenomenon of passing off inferior products as good ones, and cannot specifically identify the authenticity of Yinjiaping honey.

[0005] Therefore, it is necessary to establish a method for identifying Yinjiaping honey. Summary of the invention

[0006] In view of this, the present application provides a method for identifying Yinjiaping honey based on metabolomics technology, which is used to solve the problem of how to authenticate Yinjiaping honey.

[0007] In order to achieve the above technical objectives, this application adopts the following technical solutions: In a first aspect, the present application provides a method for identifying Yinjiaping honey based on metabolomics technology, comprising the following steps: UPLC-Q-TOF-MS was used to detect Yinjiaping honey and known group honey respectively, and Yinjiaping honey data and known group honey data were obtained. After preprocessing the Yinjiaping honey data and known group honey data, the compound components of Yinjiaping honey and known group honey were obtained respectively. The principal component analysis and orthogonal partial least squares discriminant analysis were performed on the compound components of Yinjiaping honey and known group honey, and then the differential metabolites of Yinjiaping honey and known group honey were screened by V+S plot; A PLS discriminant model was established based on the differential metabolites; The differential metabolites of the honey to be tested were determined and brought into the PLS discriminant model to obtain the predicted value of the honey to be tested, and then it was determined whether the honey to be tested was Yinjiaping honey based on the preset predicted value range.

[0008] Preferably, the parameters of UPLC-Q-TOF-MS for detecting Yinjiaping honey and known groups of honey are as follows: Chromatographic conditions: The chromatographic column was a reverse phase column; the mobile phase was A: 0.1% formic acid aqueous solution (V:V), and B: chromatographic grade methanol; the flow rate was 0.3 mL / min; the injection volume was 3 μL; the gradient elution was: 0 min, 5% B; 2 min, 5% B; 25 min, 100% B; 26 min, 5% B; 30 min, 5% B; Mass spectrometry conditions: spray voltages of 5500 and -4500 V in positive and negative ion modes, respectively, declustering voltages of 60 and -60 V, respectively; capillary temperature of 550 °C, sheath gas flow rate of 55 Arb, auxiliary gas flow rate of 55 Arb, retention time of 250 ms in full scan TOF MS mode, retention time of 70 ms in MS / MS mode, and scan range of m / z 50-750.

[0009] Preferably, the pretreatment process includes: peak extraction, retention time correction and peak alignment, and the compound components are identified using databases including HMDB, MassBank and PubChem; the software used for the pretreatment is Progenesis QI software.

[0010] Preferably, the conditions for screening differential metabolites in the V+S graph are | p [1]|>0.05,| p (corr)[1]|>0.5 and VIP>2.

[0011] Preferably, the method further includes screening or determining the differential metabolites as characteristic markers.

[0012] Preferably, the step of determining or screening the differential metabolites as characteristic markers is as follows: using Hiplot to draw a heat map of the differential metabolites and screening the corresponding potential markers from the differential metabolites, and then using SPSS 27.0 software to draw ROC curves for the potential markers, and calculate AUC, and select potential markers with AUC>0.7 as characteristic markers.

[0013] Preferably, the steps of establishing the PLS discriminant model are: selecting some samples of Yinjiaping honey and a known group of honey as training sets, assigning the categorical variables of the Yinjiaping honey samples and the known group of honey samples in the training set to 1 and 0 respectively, taking the differential metabolites of each honey sample in the training set as independent variables, and taking the categorical variables as dependent variables, to establish a PLS discriminant model.

[0014] Preferably, the step of determining whether the honey to be tested is Yinjiaping honey according to the predicted value range is: if 0.5<predicted value<1.5, the honey to be tested is Yinjiaping honey; if -0.5<predicted value<0.5, the honey to be tested is not Yinjiaping honey.

[0015] Preferably, the known group of honey includes one or more of linden honey, acacia honey, vitex honey, sweet potato honey, jujube honey, tallow tree honey, rubber honey, and rapeseed honey.

[0016] In a second aspect, the present application provides a Yinjiaping honey characteristic marker, which includes kynurenic acid, abscisic acid, D - One or more of pantothenic acid, sinapinic acid, sebacic acid, kaempferol glucuronide, 3,4,5-trimethoxycinnamic acid, 4-hydroxybenzoic acid, 4-hydroxyquinoline, alanine-phenylalanine.

[0017] The beneficial effects of the present application are as follows: the present invention is based on the non-targeted metabolomics technology of ultra-high performance liquid chromatography tandem quadrupole time-of-flight mass spectrometry (UPLC-Q-TOF-MS), screens the differential metabolites of Yinjiaping honey, and evaluates the working performance of potential markers through the receiver operating characteristic (ROC) curve, determines the characteristic markers of Yinjiaping honey, and simultaneously uses the characteristic markers to establish a partial least squares (PLS) discriminant model of Yinjiaping honey, and effectively identifies the authenticity of Yinjiaping honey through the established discriminant model. The identification method established in the present application is simple to operate, has few steps, is easy to use, has high repeatability, low detection cost, fast analysis speed, and high accuracy; The method for screening the characteristic markers of Yinjiaping honey of the present invention relies on high-throughput, high-resolution and high-precision modern analytical instruments to comprehensively and accurately analyze the components of Yinjiaping honey, and the obtained characteristic markers are used to accurately and reliably identify Yinjiaping honey. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1PCA and OPLS-DA analysis of honey samples; Figure 1 A and 1B are PCA diagrams in positive and negative ion modes, respectively. Figure 1 C and Figure 1 D are OPLS-DA diagrams in positive and negative ion modes, respectively; Figure 1 E and Figure 1 F are OPLS-DA permutation test diagrams in positive and negative ion modes, respectively; Figure 2 is a V+S diagram; Figure 2 A. Figure 2 B is the V+S diagram in positive ion mode and negative ion mode, respectively; Figure 3 This is the heat map of differential metabolites of Yinjiaping honey; Figure 4 This is the ROC curve of the characteristic markers of Yinjiaping honey; Figure 5 This is a graph showing the relationship between the true value and the predicted value in PLS analysis. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0020] Metabolomics technology relies on high-throughput, high-resolution, and high-precision modern analytical instruments to process massive data for information extraction and result analysis. Compared with traditional physical and chemical testing and previous instrumental analysis methods, metabolomics technology has obvious advantages in food identification. Metabolomics technology can achieve high-throughput simultaneous identification of multi-target full components in food, greatly improving detection throughput and efficiency, and can screen and identify markers. It is a new generation of technical means to solve the problem of food identification.

[0021] Based on this, this application was created.

[0022] The present application provides a method for identifying Yinjiaping honey based on metabolomics technology, comprising the following steps: S1. Use UPLC-Q-TOF-MS to detect Yinjiaping honey and known groups of honey, obtain raw data, and use Progenesis QI software to preprocess the raw data and identify the compound components in each honey sample; S2. Perform principal component analysis and orthogonal partial least squares discriminant analysis on the compound components of each honey sample, and then screen the differential metabolites of each honey sample through V+S plot; S3. Establish a PLS discriminant model based on differential metabolites; S4. Determine the differential metabolites of the honey to be tested, and bring them into the PLS discriminant model to obtain the predicted value of the honey to be tested; determine whether the honey to be tested is Yinjiaping honey based on the range of the predicted values.

[0023] Step S1 is a data processing process to obtain the compound components of each honey sample.

[0024] Step S2 is multivariate statistical analysis and screening of differential metabolites, wherein principal component analysis (PCA) is an unsupervised model that reduces the dimension of data by projecting variables onto the principal component with the largest variance, thereby achieving the purpose of multivariate statistical analysis. This application uses SIMCA 14.1 software to perform principal component analysis on small molecule compounds in honey samples, and uses the scores of the first two principal components to make a scatter plot; orthogonal partial least squares discriminant analysis (OPLS-DA) is a supervised multivariate statistical analysis method that can filter irrelevant variables in the matrix, maximize the differences between different samples, and highlight the differences between groups. To avoid overfitting of the model, 200 permutation tests were performed. If Q 2 The intersection with the vertical axis is below 0, and the value of the point on the left is lower than the original value on the right, indicating that the model has good accuracy and is not overfitted; the V+S graph is used to screen the differential metabolites between different groups of samples. The V+S graph combines the variable projection importance (VIP) graph and the S-plot graph. The higher the VIP value, the more important the variable; the farther the variable of the S-plot graph is from the center, the higher the content of the variable. In addition, the correlation coefficient ( p (corr)) represents the correlation between the principal component and the variable, and the covariance ( p ) represents the correlation between the principal component and the variable.

[0025] Step S3: In order to verify the role of Yinjiaping honey characteristic markers in discriminating unknown samples, the PLS discriminant model of Yinjiaping honey was established for a randomly selected portion of samples (e.g., 2 / 3) using SIMCA14.1 software, and the remaining independent honey samples (e.g., 1 / 3) were classified to evaluate the identification accuracy of the model.

[0026] Step S4 brings the differential metabolites of the honey to be tested into the PLS discriminant model for authenticity identification.

[0027] In some embodiments, the parameters for UPLC-Q-TOF-MS detection of Yinjiaping honey and known groups of honey are as follows: chromatographic conditions: the chromatographic column is a reversed phase chromatographic column; mobile phase: A is 0.1% formic acid aqueous solution (V:V), B is chromatographic grade methanol; flow rate is 0.3 mL / min; injection volume is 3 μL; gradient elution: 0 min, 5% B; 2 min, 5% B; 25 min, 100% B; 26 min, 5% B; 30 min, 5% B; mass spectrometry conditions: positive and negative ion mode spray voltages are 5500 and -4500 V, respectively, and declustering voltages are 60 and -60 V, respectively; capillary temperature is 550°C, sheath gas flow rate is 55 Arb, auxiliary gas flow rate is 55 Arb, retention time in full scan TOF MS mode is 250 ms, retention time in MS / MS mode is 70 ms, and scanning range is m / z 50-750.

[0028] In some embodiments, the pre-processing process includes: peak extraction, retention time correction and peak alignment, and the compound components are identified using databases including HMDB, MassBank and PubChem.

[0029] In some embodiments, the conditions for V+S graph screening for differential metabolites are | p [1]|>0.05,| p (corr)[1]|>0.5 and VIP>2.

[0030] The V+S graph combines the variable projection importance (VIP) graph and the S-plot graph. The higher the VIP value, the more important the variable is; the farther the variable is from the center of the S-plot graph, the higher the content of the variable. In addition, the correlation coefficient ( p (corr)) represents the correlation between the principal component and the variable, and the covariance ( p ) represents the correlation between the principal component and the variable.

[0031] In some embodiments, the method further comprises determining or screening the differential metabolites as characteristic markers.

[0032] In some embodiments, the step of determining or screening the differential metabolites as characteristic markers is as follows: using Hiplot to draw a heat map of the differential metabolites and screening the corresponding potential markers from the differential metabolites, and then using SPSS27.0 software to draw an ROC curve for the potential markers, and calculate the AUC, and select the potential markers with AUC>0.7 as the characteristic markers.

[0033] In this embodiment, it is a further verification or screening step of differential metabolites, which includes: drawing a heat map based on the differential metabolites, and then drawing a receiver operating characteristic curve based on the markers in the heat map, and calculating the area under the curve, and judging whether the differential metabolites can be used as characteristic markers of Yinjiaping honey by the area under the curve. Among them, the heat map maps each value in the data matrix into a color display according to a certain rule, uses color changes to visualize the data, intuitively presents the global content changes of multiple samples, and can also show the relationship between the contents of multiple samples. The potential markers of Yinjiaping honey are mined through the heat map; the receiver operating characteristic (ROC) curve is widely considered to be the most objective and statistically effective method for performance evaluation in biomarkers. The area under the curve (AUC) can be used to evaluate the effectiveness of the model. The closer the AUC is to 1, the stronger the discriminative ability of the model. After screening out the potential markers, the discriminative ability of the potential markers is further evaluated based on the AUC. The larger the AUC of the marker, the greater the possibility that the model classification is correct, and it can be used as a marker. SPSS 27.0 software is used to draw the ROC curve for the potential markers screened in the heat map, and the area under the curve is calculated. If the AUC of the potential marker is greater than 0.7, it can be used as a characteristic marker of Yinjiaping honey.

[0034] In some embodiments, the steps of establishing a PLS discriminant model are: selecting some samples of Yinjiaping honey and a known group of honey as training sets, assigning categorical variables of the Yinjiaping honey samples and the known group of honey samples in the training set to 1 and 0 respectively, taking the differential metabolites of each honey sample in the training set as independent variables, and taking the categorical variables as dependent variables, and establishing a PLS discriminant model.

[0035] In some embodiments, the step of determining whether the honey to be tested is Yinjiaping honey according to the predicted value range is: if 0.5<predicted value<1.5, the honey to be tested is Yinjiaping honey; if -0.5<predicted value<0.5, the honey to be tested is not Yinjiaping honey.

[0036] In some embodiments, the known group of honeys includes one or more of linden honey, acacia honey, vitex honey, sweet potato honey, jujube honey, tallow tree honey, rubber honey, and rapeseed honey.

[0037] The present application provides a Yinjiaping honey characteristic marker, which includes kynurenic acid, abscisic acid, D - One or more of pantothenic acid, sinapinic acid, sebacic acid, kaempferol glucuronide, 3,4,5-trimethoxycinnamic acid, 4-hydroxybenzoic acid, 4-hydroxyquinoline, alanine-phenylalanine.

[0038] The present solution is further described below through specific embodiments.

[0039] Source of raw materials: Yinjiaping honey: Yinjiaping honey is collected from Yinjiaping, grouped and numbered Y1-Y51; Known group honey: Commercially available non-Yinjiaping honey is grouped and numbered DZ1-DZ52.

[0040] Reagents and equipment: methanol (chromatographic grade, Thermo Fisher Scientific (China) Co., Ltd.); formic acid (chromatographic grade, Thermo Fisher Scientific (China) Co., Ltd.); pipette (Thermo Fisher Scientific (China) Co., Ltd.); analytical balance (Mettler Toledo Instruments (Shanghai) Co., Ltd.); Milli-Q ultrapure water preparation system (Wuhan Beilaimei Biotechnology Co., Ltd.); vortex oscillator (Shanghai Chuyue Electromechanical Technology Co., Ltd.); Triple TOF 5600 high-resolution time-of-flight mass spectrometer (AB Sciex, USA).

[0041] Example 1 A method for identifying Yinjiaping honey based on metabolomics technology, comprising the following steps: S1. Accurately weigh 5 g of honey samples of each number into a 50 mL centrifuge tube, add 20 mL of pH 2 acidified water to dissolve, vortex for 2 min, mix thoroughly, activate the Oasis HLB solid phase extraction column with 6 mL of pure methanol and 6 mL of pH 2 acidified water, pass the honey sample solution through the solid phase extraction column, remove the remaining sugars and unadsorbed compounds with 6 mL of pH 2 acidified water, and then collect it with pure methanol. The collected solution is re-dissolved with methanol, filtered through a 0.45 μm organic filter membrane and placed in a 2 mL brown liquid phase bottle for UPLC-Q-TOF-MS non-targeted metabolomics detection; take 10 μL of solution from each sample and place it in 10 mL centrifuge tube, mixed evenly to prepare quality control (QC) samples, before sample sequence analysis, first two injections of blank sample to clean the system, then a needle of QC sample to balance the system, during the entire analysis process, a needle of QC sample and a needle of blank sample were injected every 10 honey samples to monitor the stability of the instrument; wherein, chromatographic conditions: the chromatographic column is a reverse chromatographic column; mobile phase: A is 0.1% formic acid aqueous solution (V:V), B is chromatographic grade methanol; flow rate is 0.3 mL / min; injection volume is 3 μL; gradient elution: 0 min, 5% B; 2 min, 5% B; 25 min, 100% B; 26 min, 5% B; 30 min, 5% B; mass spectrometry conditions: positive and negative ion mode spray voltages are 5500 and -4500 V, respectively, and declustering voltages are 60 and -60 V, respectively; capillary temperature is 550 ° C, sheath gas flow rate is 55 Arb, auxiliary gas flow rate is 55 Arb, and the residence time in full scan TOF MS mode is 250 ms, the retention time in MS / MS mode was 70 ms, and the scan range was m / z50-750; After the above process, UPLC-Q-TOF-MS Yinjiaping honey and known group honey were detected to obtain raw data, and the UPLC-Q-TOF-MS raw data of the obtained honey samples were imported into Progenesis QI software for preprocessing. The preprocessing process included peak extraction, retention time correction, peak alignment, etc., and the compound components in each honey sample were identified using databases such as HMDB, MassBank, and PubChem; S2. SIMCA 14.1 was used to perform principal component analysis on the small molecule compounds in 103 honey samples. The scores of the first two principal components were used to make a scatter plot. The PCA analysis results based on the first two principal components are shown in Figure 1 ( Figure 1 A and 1B are PCA diagrams in positive and negative ion modes, respectively. Figure 1 C and Figure 1 D are OPLS-DA diagrams in positive and negative ion modes, respectively; Figure 1 E and Figure 1 F are the OPLS-DA permutation test score diagrams in positive and negative ion modes, respectively). Figure 1 A and Figure 1 B shows that the Yinjiaping honey group and the known group of honey samples can be clearly separated by the first principal component; orthogonal partial least squares discriminant analysis (OPLS-DA) is used to filter irrelevant variables in the matrix to maximize the differences between different samples and highlight the differences between groups. The OPLS-DA results show that there are significant differences in the chemical composition between the two groups. In the positive ion mode, the model can explain 98.5% (R 2 The difference between Figure 1 C); In negative ion mode, the model can explain 98.6% (R 2 The difference between ( Figure 1 D), and Q in positive and negative ion modes 2 Both are close to 1, indicating that the OPLS-DA model can effectively distinguish the two groups of samples. In order to avoid overfitting of the model, the permutation test (n = 200) was used to test the model. The results showed that R 2 =(0.0,0.0594),Q 2 =(0.0,-0.253)( Figure 1 E); in negative ion mode, R 2 =(0.0,0.138),Q 2 =(0.0,-0.274)( Figure 1 F), and the rightmost original R 2 and Q 2 The values ​​of are greater than the values ​​of the scattered points after the permutation test on the left, and Q 2The intercept values ​​are all lower than 0, indicating that the model is not overfitting; the V+S graph is then used to screen the differential metabolites between different groups of samples. The screening conditions for differential metabolites are: | p [1]|>0.05,| p (corr)[1]|>0.5,VIP>2, the results are as follows Figure 2 As shown in the figure, the V+S graph combines the variable projection importance (VIP) graph and the S-plot graph. The higher the VIP value, the more important the variable is; the farther the variable is from the center of the S-plot graph, the higher the content of the variable. In addition, the correlation coefficient ( p (corr)) represents the correlation between the principal component and the variable, and the covariance ( p ) represents the correlation between the principal component and the variable; the differential metabolites of Yinjiaping honey were analyzed, including kynurenic acid, abscisic acid, D -Pantothenic acid, sinapic acid, sebacic acid, kaempferol glucuronide, 3,4,5-trimethoxycinnamic acid, 4-hydroxybenzoic acid, 4-hydroxyquinoline, alanine-phenylalanine, and the differential metabolites and related parameters are shown in Table 1 .

[0042] Table 1 Differential metabolites and related parameters of Yinjiaping honey

[0043] The differential metabolites of Yinjiaping honey were further screened: the differential metabolites were plotted using the Hiplot website to draw a heat map, and the corresponding potential markers were screened from the differential metabolites. The results are as follows Figure 3 As shown in the figure, the distribution of differential metabolites between Yinjiaping honey and known group honey samples can be observed more intuitively. The results of the heat map show that the honey samples are divided into two categories, corresponding to Yinjiaping honey and known group honey respectively. SPSS 27.0 software is then used to draw the ROC curve for potential markers. The results are shown in Figure 4 As shown, the AUC was calculated, and potential markers with AUC>0.7 were selected as characteristic markers; kynurenic acid, abscisic acid, D The ROC curve areas of the 10 potential markers of Yinjiaping honey, including pantothenic acid, sinapic acid, sebacic acid, kaempferol glucuronide, 3,4,5-trimethoxycinnamic acid, 4-hydroxybenzoic acid, 4-hydroxyquinoline, and alanine-phenylalanine, were 0.998, 0.975, 0.995, 0.945, 0.982, 0.996, 0.982, 0.988, 0.968, and 0.990, respectively. The ROC curve analysis results are shown in the figure. Figure 4 As shown, the curves of the 10 compounds were all above the reference line, and the AUCs were all > 0.90, indicating that these 10 compounds can be used as characteristic markers of Yinjiaping honey; S3. Randomly select Yinjiaping honey and 2 / 3 of the known group honey samples (34 and 35 samples in the Yinjiaping honey group and the known group, respectively) as the training set, and the remaining 1 / 3 samples (17 samples) as the validation set. Assign 1 and 0 to the categorical variables of the corresponding honey samples in the Yinjiaping honey group and the known group in the training set, respectively. Take the differential metabolites of each honey sample in the training set as the independent variable and the categorical variable as the dependent variable, perform regression analysis, and establish a PLS discriminant model of data and categorical variables. The steps to determine whether the honey to be tested is Yinjiaping honey according to the predicted value range are as follows: if 0.5<predicted value<1.5, the honey to be tested is Yinjiaping honey; if -0.5<predicted value<0.5, the honey to be tested is not Yinjiaping honey. The results are as follows: Figure 5 A shows the correlation coefficient R between the true value and the predicted value of the categorical variable in the PLS discriminant model 2 The estimated root mean square error (RMSEE) is 0.111833, and the cross-validation root mean square error (RMSEcv) is 0.136015, indicating that the model has good discrimination accuracy. The straight line in the figure is the regression line between the model training set and the actual classification. The predicted value of the classification variable of the real group sample should be close to 1 and within the range of 0.5~1.5; the predicted value of the classification variable of the known group sample should be close to 0 and between -0.5~0.5. In order to verify the classification ability of the PLS discriminant model, the model is used to discriminate and classify 17 samples (unknown samples) in the validation set that did not participate in the model establishment. The model discrimination ability is represented by the root mean square of prediction error (RMSEP), as shown in Figure 5 As shown in Figure B, the RMSEP is 0.132563, and the honey samples in different groups can be clearly distinguished. The 17 samples in the upper right corner represent that the grouping variable is 1, and the 17 samples in the lower left corner represent that the grouping variable is 0, which just corresponds to the groups they belong to. They are all correctly discriminated, and the identification accuracy is 100%, indicating that the constructed model is effective, indicating that the PLS discriminant model established based on the characteristic markers of Yinjiaping honey can effectively identify whether the honey sample belongs to Yinjiaping honey.

[0044] The present invention is based on the non-targeted metabolomics technology of ultra-performance liquid chromatography tandem quadrupole time-of-flight mass spectrometry (UPLC-Q-TOF-MS), screens the differential metabolites of Yinjiaping honey, evaluates the working performance of potential markers through the receiver operating characteristic (ROC) curve, determines the characteristic markers of Yinjiaping honey, and simultaneously uses the characteristic markers to establish a partial least squares (PLS) discriminant model of Yinjiaping honey. The authenticity of Yinjiaping honey is effectively identified through the established discriminant model. The identification method established in the present application is simple to operate, has few steps, is easy to use, has high repeatability, low detection cost, fast analysis speed, and high accuracy. The method for screening the characteristic markers of Yinjiaping honey of the present invention relies on high-throughput, high-resolution and high-precision modern analytical instruments to comprehensively and accurately analyze the components of Yinjiaping honey, and the obtained characteristic markers are used to accurately and reliably identify Yinjiaping honey.

[0045] The above are only preferred specific implementations of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for identifying Yinjiaping honey based on metabolomics technology, characterized in that: The following steps are involved: Using UPLC-Q-TOF-MS to detect Yinjiaping honey and a known group of honey, respectively, to obtain Yinjiaping honey data and a known group of honey data, and after pre-processing the Yinjiaping honey data and the known group of honey data, to obtain compound components of Yinjiaping honey and the known group of honey, respectively; The principal component analysis and orthogonal partial least squares discriminant analysis were performed on the compound components of Yinjiaping honey and known group honey, and then the differential metabolites of Yinjiaping honey and known group honey were screened by V+S plot; Establishing a PLS discriminant model according to the differential metabolites; The differential metabolites of the honey to be tested are determined and brought into the PLS discrimination model to obtain the predicted value of the honey to be tested, and then it is determined whether the honey to be tested is Yinjiaping honey according to a preset predicted value range.

2. The method for identifying Yinjiaping honey based on metabolomics technology according to claim 1, characterized in that: The parameters of the UPLC-Q-TOF-MS for detecting Yinjiaping honey and known groups of honey are as follows: Chromatographic conditions: The chromatographic column was a reverse phase column; the mobile phase was A: 0.1% formic acid in water (V:V), and B: chromatographic grade methanol; the flow rate was 0.3 mL / min; the injection volume was 3 μL; the gradient elution was: 0 min, 5% B; 2 min, 5% B; 25 min, 100% B; 26 min, 5% B; 30 min, 5% B; Mass spectrometry conditions: spray voltages of 5500 and -4500 V in positive and negative ion modes, respectively, declustering voltages of 60 and -60 V, respectively; capillary temperature of 550 °C, sheath gas flow rate of 55 Arb, auxiliary gas flow rate of 55 Arb, retention time of 250 ms in full scan TOF MS mode, retention time of 70 ms in MS / MS mode, and scan range of m / z 50-750.

3. The method for identifying Yinjiaping honey based on metabolomics technology according to claim 1, characterized in that: The pretreatment process includes: peak extraction, retention time correction and peak alignment, and the compound components are identified using databases including HMDB, MassBank and PubChem; the software used for the pretreatment is Progenesis QI software.

4. The method for identifying Yinjiaping honey based on metabolomics technology according to claim 1, characterized in that: The conditions for screening differential metabolites in the V+S graph are: p [1]|>0.05,| p (corr)[1]|>0.5 and VIP>2.

5. The method for identifying Yinjiaping honey based on metabolomics technology according to claim 1, characterized in that: The method also includes determining or screening the differential metabolites as characteristic markers.

6. The method for identifying Yinjiaping honey based on metabolomics technology according to claim 5, characterized in that: The steps of determining or screening the differential metabolites as characteristic markers are as follows: using the Hiplot website to draw a heat map of the differential metabolites and screening the corresponding potential markers from the differential metabolites, and then using SPSS 27.0 software to draw a ROC curve for the potential markers, and calculate the AUC, and select the potential markers with AUC>0.7 as characteristic markers.

7. The method for identifying Yinjiaping honey based on metabolomics technology according to claim 1, characterized in that: The steps of establishing the PLS discriminant model are as follows: selecting some samples of the Yinjiaping honey and the known group of honey as training sets, assigning the categorical variables of the Yinjiaping honey samples and the known group of honey samples in the training sets to 1 and 0 respectively, taking the differential metabolites of each honey sample in the training set as independent variables, taking the categorical variables as dependent variables, and establishing the PLS discriminant model.

8. The method for identifying Yinjiaping honey based on metabolomics technology according to claim 7, characterized in that: The step of determining whether the honey to be tested is Yinjiaping honey according to the predicted value range is: if 0.5<predicted value<1.5, the honey to be tested is Yinjiaping honey; if -0.5<predicted value<0.5, the honey to be tested is not Yinjiaping honey.

9. The method for identifying Yinjiaping honey based on metabolomics technology according to claim 1, characterized in that: The known group of honeys includes one or more of linden honey, acacia honey, vitex honey, sweet potato honey, jujube honey, tallow tree honey, rubber honey, and rapeseed honey.

10. A characteristic marker of Yinjiaping honey obtained by the method according to any one of claims 1 to 9, characterized in that: They include kynurenic acid, abscisic acid, D - One or more of pantothenic acid, sinapinic acid, sebacic acid, kaempferol glucuronide, 3,4,5-trimethoxycinnamic acid, 4-hydroxybenzoic acid, 4-hydroxyquinoline, alanine-phenylalanine.

Images