Method for judging grade of Zhenmei green tea based on chemical components and regression equation

Abstract

The invention provides a method for judging the grade of Zhenmei green tea based on chemical components and a regression equation, and belongs to the technical field of agricultural product quality detection.The method comprises the steps that data of four target chemical components in to-be-detected Zhenmei green tea is obtained, comprising 5-gallic quinic acid, epigallocatechin gallate, gallocatechin gallate and epicatechin gallate, and is prepared from 5-gallic quinic acid, epigallocatechin gallate, gallocatechin gallate and epicatechin gallate; the data of the target chemical components are input into a regression equation, a prediction grade value corresponding to the to-be-detected Zhenmei green tea is obtained through calculation, the four target chemical components are obtained through screening based on data of known different grade sample sets, and the prediction grade value of the to-be-detected Zhenmei green tea is obtained. The regression equation is obtained through ridge regression analysis training based on data and quality grades of four target chemical components of known different-grade sample sets. The method is used for solving the problem that the grade of tea is inaccurately judged through sense organs at present, and the Zhenmei green tea grade judgment method completely based on quantifiable chemical components is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of agricultural product quality detection, and particularly relates to a grade discrimination method for Zhenmei green tea based on chemical components and regression equations and application thereof. BACKGROUND

[0002] In the export trade of Zhenmei green tea, the objective and standardized identification of its quality grade is a key link to ensure fair trade and healthy development of the industry. At present, GB / T 14456.5-2016 has made clear provisions for the grading method of Zhenmei green tea. The method mainly determines the grade of Zhenmei green tea according to sensory evaluation, and comprehensively grades the tea by evaluating the sensory properties such as appearance, soup color, aroma, taste and leaf bottom.

[0003] The inherent defect of sensory evaluation is its strong subjective dependence. There may be significant deviations in the evaluation of the same batch of tea between different evaluators or even the same evaluator at different times. This instability of the evaluation results seriously restricts the accuracy of Zhenmei green tea quality control and the fairness of trade.

[0004] Since the quality of tea is essentially a manifestation of the combination of specific chemical components, developing a Zhenmei green tea grade discrimination method that completely relies on quantifiable chemical component data and eliminates subjective factors, as a supplement and improvement to the existing sensory evaluation system, has a positive significance for improving the objectivity, stability and efficiency of Zhenmei green tea grade identification. SUMMARY

[0005] In view of the shortcomings of the above-mentioned sensory evaluation technology, the purpose of the present application is to provide a Zhenmei green tea grade discrimination method based on chemical components and regression equations, aiming to solve the problem of inaccurate tea grade determination by sensory judgment.

[0006] The present application provides a Zhenmei green tea grade discrimination method based on chemical components and regression equations, comprising the following steps:

[0007] ① Number the grades of all Zhenmei green tea samples, with the grades from high to low being assigned values of 1.00 to 12.00, respectively;

[0008] ② Select at least 4 Zhenmei green tea samples of different known grades, and obtain the peak areas corresponding to 5-galloylquinic acid, epigallocatechin gallate, gallocatechin gallate and epicatechin gallate by chromatography-mass spectrometry;

[0009] ③ Establish a ridge regression prediction model by leave-one-out cross-validation method, and based on the intercept of the ridge regression model and the regression coefficients of the peak areas of the four compounds in step ②, construct a grade discrimination regression equation;

[0010] ④ Using the same method as in step ②, determine the peak areas of 5-gallicin, epigallocatechin gallate, gallatechin gallate, and epicatechin gallate in the unknown grade Zhenmei green tea sample, and substitute the peak areas into the grade discrimination regression equation obtained in step ③ to calculate the predicted grade value corresponding to the Zhenmei green tea to be tested.

[0011] The peak areas of 5-gallic quinic acid, epigallocatechin gallate, gallocatechin gallate, and epigallocatechin gallate are determined by retention time and mass-to-charge ratio.

[0012] The method for determining the retention time and mass-to-charge ratio is as follows:

[0013] A. Grind Zhenmei green tea into powder, and extract the Zhenmei green tea powder with methanol aqueous solution to obtain Zhenmei green tea extract;

[0014] B. Data acquisition of the extract was performed using chromatography-mass spectrometry.

[0015] C. Process the collected data to obtain variable data of non-volatile metabolites;

[0016] D. Based on the tea metabolite database, standards, and LC-Q-TOF-MS secondary mass spectrometry fragmentation analysis, the retention times, mass-to-charge ratios, and peak areas of 5-gallicin, epigallocatechin gallate, gallatechin gallate, and epigallocatechin gallate were determined.

[0017] Among them, in the chromatography-mass spectrometry technique,

[0018] Chromatographic liquid phase conditions: Column Acquity BEH Shield RP 18 The column was set with a flow rate of 0.25 mL / min, a detection wavelength of 278 nm, a column temperature of 40 °C, and an injection volume of 2 μL. The mobile phase consisted of 0.1% formic acid in water (phase A) and acetonitrile in acetonitrile (phase B). The elution gradient was as follows: 0–8 min: 2%–15% B; 8–14 min: 15%–30% B; 14–16 min: 30%–50% B; 16–18 min: 50%–90% B; 18–20 min: 90%–95% B; 20–22 min: 95%–2% B; 22–26 min: 2% B.

[0019] The mass spectrometry conditions are: electrospray ion source, capillary voltage 3500 V, nozzle voltage 500 V, cone voltage 75 V, capillary outlet voltage 185 V, drying gas temperature 320℃, drying gas flow rate 8 L / min, sheath gas temperature 350℃, sheath gas flow rate 11 L / min, sheath gas, auxiliary gas and collision gas are all high-purity nitrogen, collected in negative ion mode, and the first ion scanning range is 100-1500 m / z.

[0020] The preparation method of the Zhenmei green tea extract is as follows: tea leaves are ground into powder by using a grinder, 70% methanol is added, ultrasonic-assisted extraction is performed, and then centrifugation is performed to collect the supernatant; the supernatant is accurately diluted with 70% methanol, filtered through a 0.22 μm filter membrane, and the Zhenmei green tea extract is obtained.

[0021] The grade classification method of the Zhenmei green tea sample is the grade classification method of Zhenmei green tea specified in GB / T 14456.5-2016.

[0022] The regression equation is as follows:

[0023] Predicted grade = 25.890-1.003×10 –7 ×5-Galloylquinic acid-4.930×10 –8 ×EGCG-4.741×10 –7 ×GCG-1.166×10 –7 ×ECG

[0024] 5-Galloylquinic acid = the peak area of 5-galloylquinic acid;

[0025] EGCG = the peak area of epigallocatechin gallate;

[0026] GCG = the peak area of gallocatechin gallate;

[0027] ECG = the peak area of epicatechin gallate.

[0028] In step 2, preferably 4-12 Zhenmei green tea samples of different known grades are used.

[0029] Beneficial effects

[0030] According to the Zhenmei green tea grade discrimination method based on chemical components and a regression equation, the training process of the regression equation is as follows: data of four target chemical components corresponding to a known sample set of different grades are obtained; the regression coefficient and intercept between the data of the target chemical components and the known grade values are determined by ridge regression analysis; and the regression equation is constructed according to the regression coefficient and intercept; the determination coefficient (R2 ) is 0.995, the root mean square error (RMSE) of leave-one-out cross-validation is 0.168 grade units, and the mean absolute error (MAE) is 0.139, indicating that the regression equation has high prediction accuracy. Compared with the traditional grading method relying on the sensory experience of evaluators, the use of the present application can realize objective quantification of grades, eliminate subjective bias, and at the same time complete the rapid detection and analysis of batch samples in a short time, significantly improving efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0031] Figure 1 is a compound plot of the frequency selected in 100 bootstrap samples, where the dashed line represents the 70% frequency threshold;

[0032] Figure 2 is a regression equation validation result plot of Example 1.

[0033] Figure 3 is a regression equation validation result plot of a random tea factory. DETAILED DESCRIPTION

[0034] The present application provides a Zhenmei green tea grade discrimination method based on chemical composition and regression equation, and the purpose, technical scheme and effect of the present application are more clear and explicit, which is further described in detail as follows.

[0035] Example 1

[0036] This example is based on Zhenmei green tea produced in Anhui, and the Zhenmei green tea grade discrimination method is modeled through chemical composition analysis, key predictor search and regression equation, etc. The detailed process is as follows.

[0037] 1. Collection and extraction of tea samples

[0038] In order to minimize the impact of different processing differences on metabolites, we collected eight different grades of Zhenmei green tea from the same tea factory and the same batch, including six special Zhenmei and two ordinary Zhenmei. According to GB / T 14456.5-2016 "Green Tea Part 5: Mei Tea", the samples were sorted by grade from high to low, and the detailed information is shown in Table 1.

[0039] The tea leaves were ground into powder using a grinder and accurately weighed 100 mg, and transferred to a 5 mL centrifuge tube. Add 3 mL of 70% methanol to each centrifuge tube, and extract at 25°C for 10 minutes. Then centrifuge at 8000 rpm for 10 minutes, collect the supernatant. Repeat the extraction of the residue twice, and combine the supernatant to a 10 mL volumetric flask, and accurately dilute to 10 mL with 70% methanol, each sample is extracted in triplicate. Before analysis, each extraction sample is filtered through a 0.22 μm filter membrane and placed in a sample bottle.

[0040] Table 1, Detailed information of eight different grades of tea samples

[0041]

[0042] 2. Data collection

[0043] Instrument: Agilent Infinity 1290 UPLC-G6545 Q-TOF / MS (liquid chromatography tandem quadrupole time-of-flight mass spectrometry) was used.

[0044] Liquid phase conditions: Chromatographic column Acquity BEH Shield RP 18 column (2.1 x 50 mm, 1.7 μm, Waters, USA) with a flow rate of 0.25 mL / min, detection wavelength of 278 nm, column temperature of 40 °C, and injection volume of 2 μL; the mobile phase A was 0.1% formic acid water and the mobile phase B was acetonitrile; the elution gradient was as follows: 0-8 min: 2%-15% B; 8-14 min: 15%-30% B; 14-16 min: 30%-50% B; 16-18 min: 50%-90% B; 18-20 min: 90%-95% B; 20-22 min: 95%-2% B; 22-26 min: 2% B.

[0045] Mass spectrometry conditions: Electrospray ion source, capillary voltage 3500 V, nozzle voltage 500 V, cone hole voltage 75 V, capillary outlet voltage 185 V, drying gas temperature 320 °C, drying gas flow rate 8 L / min, sheath gas temperature 350 °C, sheath gas flow rate 11 L / min, sheath gas, auxiliary gas, and collision gas were all high-purity nitrogen, collected in negative ion mode, primary ion scan range: 100-1500 (m / z).

[0046] Reagents: Formic acid (mass spectrometry grade) and acetonitrile (mass spectrometry grade) were purchased from Tedia Company, USA.

[0047] 3. Data processing

[0048] The mass spectrometry data obtained by data collection for the extracted samples were processed by MS-DIAL (v 4.9) to obtain variable data of non-volatile metabolites, including retention time (min), mass-to-charge ratio (m / z), and peak area.

[0049] 4. Obtain marker differential compounds

[0050] Variable Importance in Projection (VIP) was used to screen non-volatile compounds related to the grade classification by Simca-P (v 14.1) Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA). We selected compounds with VIP > 1 as the marker compounds to distinguish the grade of Zhenmei green tea, and identified some of the differential compounds based on tea metabolite database, standard substances and LC-Q-TOF-MS secondary mass spectrum fragment analysis.

[0051] 5. Construction of grade prediction model

[0052] R language (v 4.4.1) was used to process the data of differential compounds, and caret and glmnet packages were used to build the prediction model. The modeling process was divided into two stages: key predictor screening and prediction model construction.

[0053] 5.1 Key predictor screening

[0054] Elastic net regression was used to screen compounds with predictive ability for tea grade from the initial 68 differential compounds, and 100 times of Bootstrap (self-sampling) method was used to enhance the stability of selection. The proportion of variables selected in 100 times of self-sampling was the selection frequency. As shown in Table 4, the selection frequency of compound 5-Galloylquinic acid was the highest, indicating that it had a stable contribution to the prediction of tea grade. Based on the selection frequency threshold of > 70%, 5-Galloylquinic acid, EGCG, ECG and GCG (selection frequencies were 95.0%, 84.0%, 84.0% and 74.0%, respectively) were selected as key grade predictors for subsequent ridge regression modeling. Figure 1

[0055] 5.2 Construction of prediction model

[0056] According to the tea number of Zhenmei green tea, the grades were assigned values from 1.00 to 12.00 from high to low, and the eight tea samples were numbered as shown in Table 2. Based on the peak area of the key predictors of the eight tea samples with different grades, a ridge regression prediction model was established by Leave-One-Out Cross-Validation (LOOCV). The indicators of the ridge regression model, the intercept and the regression coefficients of the compounds are shown in Table 3.

[0057] Table 2, Grade values corresponding to the tea number of Zhenmei green tea

[0058]

[0059] ​Table 3, Index, intercept and regression coefficient of the compound of the ridge regression model

[0060]

[0061] In the model, the ridge regression coefficient is negative, which indicates a positive contribution to the grade. The greater the grade prediction value obtained from the model, the lower the quality grade of the tea sample.

[0062] 6. Constructing the regression equation for grade discrimination

[0063] Based on the intercept and regression coefficient of the compound of the ridge regression model, the regression equation for grade discrimination is constructed as follows:

[0064] Predicted grade = 25.890 - 1.003 x 10 –7 x 5-Galloylquinic acid - 4.930 x 10 –8 x EGCG - 4.741 x 10 –7 x GCG - 1.166 x 10 –7 x ECG

[0065] Note: 5-Galloylquinic acid = peak area of 5-galloylquinic acid;

[0066] EGCG = peak area of epigallocatechin gallate;

[0067] GCG = peak area of gallocatechin gallate;

[0068] ECG = peak area of epicatechin gallate;

[0069] Based on the above regression equation, the training set tea samples were verified, and the verification results are shown in Table 4, which shows that the model has high prediction accuracy. Figure 2

[0070] In this embodiment, 8 tea samples are used to determine the regression equation for grade discrimination. According to the principle of linear regression equation, 4-12 tea samples can be randomly selected to obtain the regression equation for grade discrimination.

[0071] Example 2

[0072] Based on the regression equation of the tea produced in Anhui in Example 1, the inventors applied the model to 3 tea factories that randomly produced Zhenmei green tea, labeled as H factory, Z factory and S factory. The specific experimental method is as follows.

[0073] 1. Collection and extraction of tea samples

[0074] ​Three tea factories (H, Z, S) producing Zhenmei green tea were randomly selected, and several different grades of Zhenmei green tea were purchased from each tea factory, forming a total of 15 tea samples to be tested. According to the corresponding grade values in Table 2, the grade values of the tea samples to be tested were numerized, and the detailed information is shown in Table 4.

[0075] Table 4 Detailed information of different tea samples of three factories

[0076]

[0077] 2. The same method of data acquisition, data processing, obtaining the data of target objects, and predicting model construction, the indexes, intercepts and regression coefficients of compounds of the ridge regression model of each tea sample of the three factories are shown in Table 5.

[0078] Table 5 Indexes, intercepts and regression coefficients of compounds of the ridge regression model

[0079]

[0080] The above input is brought into the prediction grade regression equation of Example 1

[0081] Predicted grade = 25.890 - 1.003 x 10 –7 x 5-Galloylquinic acid - 4.930 x 10 –8 x EGCG - 4.741 x 10 –7 x GCG - 1.166 x 10 –7 x ECG

[0082] The results of the predicted grade and the actual grade are shown in Figure 3 .

[0083] The results show that the order of the predicted grade of each tea sample is consistent with the order of the actual labeled grade. It can be used to quickly predict the grade of Zhenmei green tea.

[0084] In summary, the Zhenmei green tea grade discrimination method based on chemical composition and regression equation provided by the present application clearly quantifies the quantitative contribution of four chemical components to grade prediction, and has a certain effectiveness for grade determination of Zhenmei green tea, and provides a new tool for objective quantification of tea quality.

Claims

1. A method for determining the grade of Zhenmei green tea based on chemical composition and regression equation, characterized in that, The method for determining the grade of Zhenmei green tea includes the following steps: ① The grades of all Zhenmei green tea samples were numerically assigned, with grades ranging from 1.00 to 12.00 from high to low; ② Select at least 4 different known grades of Zhenmei green tea samples and obtain the peak areas corresponding to 5-gallicin, epigallocatechin gallate, gallatechin gallate and epigallocatechin gallate by chromatography-mass spectrometry. ③ A ridge regression prediction model is established using leave-one-out cross-validation. Based on the intercept of the ridge regression model and the regression coefficients of the peak areas of the four compounds in step ②, a grade discriminant regression equation is constructed. ④ Using the same method as in step ②, determine the peak areas of 5-gallicin, epigallocatechin gallate, gallatechin gallate, and epicatechin gallate in the unknown grade Zhenmei green tea sample, and substitute the peak areas into the grade discrimination regression equation obtained in step ③ to calculate the predicted grade value corresponding to the Zhenmei green tea to be tested.

2. The method for determining the grade of Zhenmei green tea based on chemical composition and regression equation according to claim 1, characterized in that, The peak areas of 5-gallic quinic acid, epigallocatechin gallate, gallocatechin gallate, and epigallocatechin gallate were determined by retention time and mass-to-charge ratio.

3. The method for determining the grade of Zhenmei green tea based on chemical composition and regression equation according to claim 2, characterized in that, The method for determining the retention time and mass-to-charge ratio is as follows: A. Grind Zhenmei green tea into powder, and extract the Zhenmei green tea powder with methanol aqueous solution to obtain Zhenmei green tea extract; B. Data acquisition of the extract was performed using chromatography-mass spectrometry. C. Process the collected data to obtain variable data of non-volatile metabolites; D. Based on the tea metabolite database, standards, and LC-Q-TOF-MS secondary mass spectrometry fragmentation analysis, the retention times, mass-to-charge ratios, and peak areas of 5-gallicin, epigallocatechin gallate, gallatechin gallate, and epigallocatechin gallate were determined.

4. The method for determining the grade of Zhenmei green tea based on chemical composition and regression equation according to claim 3, characterized in that: In the aforementioned chromatography-mass spectrometry technique Chromatographic liquid phase conditions: Column Acquity BEH Shield RP 18 The column was set with a flow rate of 0.25 mL / min, a detection wavelength of 278 nm, a column temperature of 40 °C, and an injection volume of 2 μL. The mobile phase consisted of 0.1% formic acid in water (phase A) and acetonitrile in acetonitrile (phase B). The elution gradient was as follows: 0–8 min: 2%–15% B; 8–14 min: 15%–30% B; 14–16 min: 30%–50% B; 16–18 min: 50%–90% B; 18–20 min: 90%–95% B; 20–22 min: 95%–2% B; 22–26 min: 2% B. Mass spectrometry conditions: electrospray ionization source, capillary voltage 3500V, nozzle voltage 500V, cone voltage 75V, capillary outlet voltage 185V, drying gas temperature 320℃, drying gas flow rate 8L / min, sheath gas temperature 350℃, sheath gas flow rate 11L / min, sheath gas, auxiliary gas, and collision gas are all high-purity nitrogen, collected in negative ion mode, primary ion scan range: 100-1500m / z.

5. The method for determining the grade of Zhenmei green tea based on chemical composition and regression equation according to claim 3, characterized in that: The method for preparing the Zhenmei green tea extract is as follows: the tea leaves are ground into powder using a grinder, 70% methanol is added, and after ultrasonic-assisted extraction, the mixture is centrifuged, the supernatant is collected, the supernatant is precisely diluted with 70% methanol, and filtered through a 0.22μm filter membrane to obtain the Zhenmei green tea extract.

6. The method for determining the grade of Zhenmei green tea based on chemical composition and regression equation according to claim 1, characterized in that, The grading method for the Zhenmei green tea sample is the grading method for Zhenmei green tea specified in GB / T 14456.5-2016.

7. The method for determining the grade of Zhenmei green tea based on chemical composition and regression equation according to claim 1, characterized in that, The regression equation is as follows: Prediction level = 25.890 – 1.003 × 10 –7 ×5-Galloylquinic acid–4.930×10 –8 ×EGCG–4.741×10 –7 ×GCG–1.166×10 –7 ×ECG Wherein: 5-Galloylquinic acid = the peak area of ​​5-galloylquinic acid; EGCG = Peak area of ​​catechin gallate; GCG = Peak area of ​​gallocatechin gallate; ECG = Peak area of ​​epicatechin gallate.

8. The method for determining the grade of Zhenmei green tea based on chemical composition and regression equation according to claim 1, characterized in that, In step ②, 4 to 12 different known grades of Zhenmei green tea samples are preferred.

Images