A method for simultaneously detecting the origin and glycogen content of oysters based on ATR-FTIR

Abstract

The present invention discloses a method for simultaneously detecting the origin and glycogen content of oysters based on ATR-FTIR, comprising: step one, detecting the ATR-FTIR data and glycogen content of oyster tissue dry powder from different origins; step two, importing spectral analysis software, performing second-order derivative pretreatment to obtain pre-processed spectral data of the origin; performing multivariate scattering correction and second-order derivative pretreatment to obtain pre-processed spectral data of glycogen content; step three, establishing an origin discrimination model based on the pre-processed spectral data of the origin and the origin; establishing a glycogen content quantitative model based on the pre-processed spectral data of the glycogen content and the glycogen content; step four, inputting the pre-processed spectral data of the origin of the sample to be tested and the pre-processed spectral data of glycogen into the origin discrimination model and the glycogen content quantitative model, respectively, to obtain the origin and glycogen content. The present invention has the advantages of only detecting ATR-FTIR data, and can quickly and accurately discriminate the origin and glycogen content of oyster samples.

Description

Technical Field

[0001] The present invention relates to the technical field of oyster detection, and more specifically, to a method for simultaneously detecting the origin and glycogen content of oysters based on ATR-FTIR. Background Art

[0002] Oysters, also known as oysters, sea oysters, oyster roe, oyster clams, oysters, ancient oysters, and oysters, belong to the family Ostridia and Ostreae of the order Pectinata of the class Bivalvia of the phylum Mollusca. They are the most important cultured shellfish in the world and one of the four major cultured shellfish in my country. As a popular seafood, oysters have attracted much attention due to their high nutritional content and delicious taste. In addition to being a precious food resource, oysters have been widely used as marine Chinese medicinal materials for centuries. Its meat is considered effective in treating chronic diseases and health care.

[0003] Geographical indication products refer to products that are "produced in a specific region using raw materials produced in a specific region according to traditional techniques, whose quality, characteristics or reputation essentially depends on the geographical characteristics of their region of origin and are named after the region of origin after review and approval in accordance with these regulations." At present, Rushan oysters and Zhanjiang oysters have been certified as national geographical indication products due to their unique regional characteristics and quality characteristics. The oyster origin traceability certification system is related to food safety, consumer rights protection, and the development of oyster aquaculture. At present, my country's seafood traceability research and traceability system are still relatively backward, and it is urgent to establish a reliable and accurate seafood certification method.

[0004] Glycogen is the main form of glucose reserve in bacteria, fungi, mammalian cells and bivalve mollusks, and is considered a key molecular contributor to flavor and quality. Oysters with higher glycogen content generally taste better. Oyster glycogen has been reported to have antioxidant, hepatoprotective, anti-tumor, and anti-inflammatory biological activities. Currently, colorimetric methods are mostly used to determine glycogen content, which often require the use of hazardous and toxic solvents and longer sample preparation and analysis procedures. Therefore, a simple and green glycogen analysis method is needed to improve the efficiency of glycogen detection in oysters. Summary of the invention

[0005] An object of the present invention is to solve at least the above problems and to provide at least the advantages which will be described hereinafter.

[0006] In order to achieve these purposes and other advantages according to the present invention, a method for simultaneously detecting the origin and glycogen content of oysters based on ATR-FTIR is provided, comprising the following steps:

[0007] Step 1: Detect the ATR-FTIR data and glycogen content of oyster tissue dry powder from different origins;

[0008] Step 2: Import the ATR-FTIR data in step 1 into the spectrum analysis software, and perform second-order derivative preprocessing to obtain preprocessed spectrum data of the origin;

[0009] The ATR-FTIR data in step 1 is imported into the spectral analysis software, and multivariate scattering correction and second-order derivative preprocessing are performed to obtain preprocessed spectral data of glycogen content;

[0010] Step 3: Establishing a partial least squares discriminant analysis model of the origin based on the pre-processed spectral data and the origin, i.e., the origin discriminant model;

[0011] A synergistic interval partial least squares regression model was established based on the pre-processed spectral data and glycogen content data, thus obtaining a quantitative model for glycogen content.

[0012] Step 4: Use the methods of step 1 and step 2 to obtain the pre-processed spectral data of the origin of the oyster tissue dry powder and the pre-processed spectral data of glycogen of the sample to be tested, and input them into the origin discrimination model and the glycogen content quantification model respectively, and output the corresponding origin and glycogen content of the sample to be tested respectively.

[0013] Preferably, the origin of the oyster tissue includes Qinzhou City, Guangxi Zhuang Autonomous Region, Zhanjiang City, Guangdong Province, Dongshan County, Fujian Province, Cangnan County, Zhejiang Province, Qinhuangdao City, Hebei Province, the Yellow Sea area of ​​Dalian City, Liaoning Province, the Bohai Sea area of ​​Dalian City, Liaoning Province, and Rushan City, Shandong Province.

[0014] Preferably, the spectral range of ATR-FTIR is 650-4000cm -1 .

[0015] Preferably, the method for establishing the origin discrimination model specifically includes:

[0016] The Kennard-Stone algorithm is used to divide the pre-processed spectral data and the origin of the oyster tissue dry powder from each origin into a training set and a validation set, wherein the training set is 3 / 4 of the total number of samples, the training set is used to establish a partial least squares discriminant analysis model of the origin, and the validation set is used to verify the partial least squares discriminant analysis model of the origin;

[0017] The origin preprocessing data obtained in step 2 is used as an input variable, and the origin variable is used as an output variable to be imported into the statistical analysis software for normalization processing to obtain the origin discrimination model.

[0018] Preferably, the method for establishing the glycogen content quantitative model specifically comprises:

[0019] The pre-processed spectral data and glycogen content data of oyster tissue dry powder from various origins were divided into a training set and a validation set using the Kennard-Stone algorithm, wherein the training set was 3 / 4 of the total number of samples, the training set was used to establish a synergistic interval partial least squares regression model, and the validation set was used to validate the synergistic interval partial least squares regression model;

[0020] The pre-processed spectral data of glycogen content is used as an input variable, and the glycogen content is used as an output variable to import into statistical analysis software to establish the glycogen content quantitative model.

[0021] Preferably, the method for preparing oyster tissue dry powder comprises: washing the oyster sample, dissecting the complete oyster tissue, freeze-drying, and grinding until the particle size of the oyster tissue is less than 100 mesh.

[0022] The present invention includes at least the following beneficial effects: a green, environmentally friendly and simple method for geographical authentication and glycogen detection of oysters is provided to ensure the quality of oysters. The method also solves the problem of difficulty in identifying the origin of oysters and time-consuming and expensive traditional glycogen detection. The detection method of the present invention has the advantages of high accuracy, small sample consumption, and no need for standard products. The method has an accuracy of 100% in determining the origin, a correlation coefficient of glycogen content prediction of 0.96, and a predicted relative analytical error RPD of 3.58.

[0023] Other advantages, objectives and features of the present invention will be embodied in part through the following description, and in part will be understood by those skilled in the art through study and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 ATR-FTIR spectra of all oyster samples;

[0025] Figure 2 Representative ATR-FTIR spectra of oysters from different origins;

[0026] Figure 3 It is the three-dimensional score graph of the PLS-DA model after second-order derivative preprocessing;

[0027] Figure 4 200 permutation tests for the PLS-DA model after second-order derivative preprocessing;

[0028] Figure 5 Interval selection results for SI-PLS model;

[0029] Figure 6 This is the prediction result of glycogen content using SI-PLS model. DETAILED DESCRIPTION

[0030] The present invention is further described in detail below in conjunction with the accompanying drawings so that those skilled in the art can implement the invention with reference to the description.

[0031] It should be noted that the experimental methods described in the following embodiments are conventional methods unless otherwise specified, and the reagents and materials can be obtained from commercial channels unless otherwise specified.

[0032] 1. Instruments and Equipment

[0033] ATR-FTIR spectra were acquired using a Nicolet iS20 Fourier transform infrared spectrometer from Thermo Fisher Scientific, USA, equipped with a ZnSe ATR crystal;

[0034] LC-12N-50B vacuum freeze dryer (Shanghai Lichen Bangxi Instrument Technology Co., Ltd.).

[0035] The original spectra of the samples were collected by OMNIC9.0, and TQ Analyst EZ Edition9.1 software was used to preprocess the spectra.

[0036] 2. Samples

[0037] The experiment collected 128 samples of oysters from 8 different origins. The origins were Zhanjiang City, Guangdong (n=16), Cangnan County, Zhejiang (n=17), Rushan City, Shandong (n=12), Qinzhou City, Guangxi (n=16), Qinhuangdao City, Hebei Province (n=17), Yellow Sea Waters of Dalian City, Liaoning Province (n=16), Dongshan County, Fujian Province (n=18), Bohai Sea Waters of Dalian City, Liaoning Province (n=16), and the number of samples in each group was represented by n. Wash the oyster samples from different origins to be tested, dissect the complete oyster tissue and freeze-dry it for 48 hours, grind the dried oysters evenly to obtain oyster dry powder, and finally pass through a 100-mesh standard sieve to obtain the sample. All samples were finally stored at -20°C for use; the glycogen reference value of the oyster sample was obtained by testing with a glycogen detection kit (Beijing Solebold Technology Co., Ltd.).

[0038] 3. Test methods and test results

[0039] 3.1 Infrared spectrum acquisition

[0040] Each sample was tested at room temperature with 16 scans and a resolution of 4 cm -1 , scanning range is 650~4000cm -1 The oyster sample was pressed against the ZnSe crystal and the torque knob was used to maintain the same pressure during the test. After each oyster sample was tested, the ZnSe crystal was cleaned with anhydrous ethanol. Each sample was measured three times and the average value was taken.

[0041] 3.2ATR-FTIR spectrum analysis and characterization

[0042] like Figure 2 As shown, 3000~2800cm -1 The peak at 1745cm is the CH vibration region, which is mainly caused by lipids and proteins. -1 The stretching vibration of the ester carbonyl (C=O) at 1644 cm is caused by the ester functional group in the lipid. -1 The peak is related to 1546cm -1 The peak at 1200~1000cm belongs to the vibration absorption of CN and NH functional groups (amide II band). The peaks of amide I band and amide II band are mainly composed of various amino acids in proteins. -1 The peaks are mainly related to CC stretching vibration, COC stretching vibration, PO-2 symmetric stretching vibration and SO3 symmetric stretching vibration, and are mainly related to substances such as proteins, phospholipids, glycogen and nucleic acids.

[0043] 3.3 Spectral preprocessing method

[0044] The original spectrum of oysters is interfered by background and noise information. In order to maximize the variability of the required information between samples and minimize irrelevant information, the original spectrum data is preprocessed to reduce noise information. Different preprocessing methods will lead to different results. The study compared different signal preprocessing methods to improve the performance of classification models and quantitative models. These processing methods include multivariate scattering correction, first-order derivative, second-order derivative preprocessing, multivariate scattering correction + first-order derivative, and multivariate scattering correction + second-order derivative.

[0045] The qualitative model fitting process uses Q 2 Come and evaluate, Q 2 It indicates the predictive ability of the classification model. The closer the value is to 1, the better the model is, and the lower the value is, the worse the fitting accuracy of the model is. The quantitative model uses the internal cross-validation relative analysis error (RPD CV ), external validation relative analytical error (RPD CV ) to evaluate the effect of preprocessing methods on the performance of partial least squares regression (PLSR) models. RPD>3 represents high model performance and good robustness. The results are shown in Tables 1 and 2. The partial least squares discriminant model Q 2 The value is the highest, so the second-order derivative is used as a preprocessing method for the discriminant model. RPD of multivariate scatter correction + second-order derivative in the quantitative model CV and RPD V All of them were greater than 3, so multivariate scatter correction + second-order derivative was used as the preprocessing method for the glycogen quantitative model.

[0046] Table 1 Parameters of partial least squares discriminant model (PLS-DA model) based on different preprocessing methods

[0047] serial number Preprocessing methods <![CDATA[R 2 X]]> <![CDATA[R 2 And]]> <![CDATA[Q 2 ]]> 1 Original spectrum 0.994 0.666 0.524 2 Multivariate Scattering Correction 0.997 0.821 0.628 3 Multivariate scatter correction + first derivative 0.844 0.983 0.944 4 Multivariate scatter correction + second derivative 0.727 0.984 0.949 5 First Derivative 0.888 0.986 0.95 6 Second Derivative 0.72 0.986 0.96

[0048] Table 2 Effects of different preprocessing methods on the partial least squares regression model (SI-PLS)

[0049]

[0050] Rcv (training set cross-validation correlation coefficient); RMSEC (training set cross-validation root mean square error); RPDcv (training set cross-validation relative analysis error); Rv (validation set correlation coefficient); RMSEP (validation set root mean square error); RPDv (validation set relative analysis error)

[0051] 3.4 Partial least squares discriminant analysis model establishment

[0052] The Kennard-Stone algorithm was used to divide the spectral data of oyster powder from various origins into a training set and a validation set. The training set was used to establish a discriminant model of origin, and the validation set was used to verify the accuracy of the above model. The raw data was preprocessed with the second-order derivative and imported into the SIMCA-P+14.1 software. The number of principal components for optimizing modeling was verified using a 7-fold cross-validation. 200 permutation tests were used to evaluate whether the partial least squares discriminant analysis model was overfitting. The permutation test results are shown in Figure 2. Figure 4 As shown, all left R 2 and Q 2 All points are below the original point, and the intercept Q 2 <0, indicating that the constructed model is reliable and has a low risk of overfitting. Figure 3 The three-dimensional score diagram of the established PLS-DA model is Figure 3 It can be seen that oyster samples from different origins showed an obvious separation trend.

[0053] 3.5 Origin determination

[0054] The validation set data was imported into the partial least squares discriminant analysis model to discriminate the origin of oysters. The classification results of the support vector machine model under different pretreatments are shown in Table 3. The number of false positives was 0 and the classification accuracy was 100%. This shows that the ATR-FTIR combined with the partial least squares discriminant analysis model can accurately identify the origin of oysters.

[0055] Table 3 Classification results of the validation set of the PLS-DA model based on the second-order derivative

[0056] quantity Accuracy A B C D E F G H No group A 6 100％ 6 0 0 0 0 0 0 0 0 B 0 100％ 0 0 0 0 0 0 0 0 0 C 3 100％ 0 0 3 0 0 0 0 0 0 D 8 100％ 0 0 0 8 0 0 0 0 0 E 7 100％ 0 0 0 0 7 0 0 0 0 F 5 100％ 0 0 0 0 0 5 0 0 0 G 1 100％ 0 0 0 0 0 0 1 0 0 H 2 100％ 0 0 0 0 0 0 0 2 0 No group 0 0 0 0 0 0 0 0 0 0 total 32 100.00％ 4 3 3 4 4 4 6 4 0

[0057] Note: A is Dongshan County, Fujian, B is Cangnan County, Zhejiang, C is Dalian City, Liaoning (Bohai Sea), D is Dalian City, Liaoning (Yellow Sea), E is Qinzhou City, Guangxi, F is Qinhuangdao City, Hebei, G is Rushan City, Shandong, and H is Zhanjiang City, Liaoning, Guangdong.

[0058] 3.6 Establishment of collaborative interval partial least squares regression model.

[0059] The Kennard-Stone algorithm was used to divide the spectral data and glycogen content of oyster powder from various origins into a training set and a validation set. The training set was used to establish the model, and the validation set was used to verify the model. The full spectrum was divided into spectral sub-intervals of widths of 10, 15, 20, 25, and 30, and 2, 3, and 4 spectral sub-intervals were combined. The combined spectral data was used as the input variable, and the glycogen content was used as the output variable to establish a collaborative partial least squares regression model (PLSR).

[0060] 3.7 Oyster glycogen prediction

[0061] The validation set data was imported into the collaborative partial least squares regression model. The prediction results of the collaborative partial least squares regression model for the glycogen content of the validation set are shown in Table 4. When the spectrum was divided into 25 intervals and the interval combination was 4, the prediction root mean square error and relative analysis error were the smallest, which were 0.0174 and 3.58, respectively. At this time, the intervals selected by the model are as follows Figure 5 As shown, they are 2794~2928cm -1 1320~1588cm -1 、1052~1186cm -1 The prediction results of the collaborative interval partial least squares regression model are as follows: Figure 6 As shown in the figure, the predicted value and the reference value (measured value) have a high correlation, and the prediction correlation coefficient is 0.9565. The above results show that the synergistic interval partial least squares regression model can well predict the content of glycogen in oysters with high accuracy.

[0062] Table 4 The influence of interval division and combination on the prediction effect of PLSR model

[0063] Divide the interval Interval combination method RMSEP RPDV 10 2 0.018 3.46 10 3 0.0178 3.50 10 4 0.0182 3.43 15 2 0.0214 2.91 15 3 0.0215 2.90 15 4 0.0209 2.98 20 2 0.0184 3.39 20 3 0.02 3.12 20 4 0.0199 3.13 25 2 0.0184 3.39 25 3 0.0185 3.37 25 4 0.0174 3.58 30 2 0.0207 3.01 30 3 0.0209 2.98 30 4 0.0207 3.01

[0064] Although the embodiments of the present invention have been disclosed as above, they are not limited to the applications listed in the specification and the implementation modes, and they can be fully applied to various fields suitable for the present invention. For those familiar with the art, additional modifications can be easily implemented. Therefore, without departing from the general concept defined by the claims and the scope of equivalents, the present invention is not limited to the specific details and the illustrations shown and described herein.

Claims

1. A method for simultaneously detecting the origin and glycogen content of oysters based on ATR-FTIR, characterized in that: The following steps are involved: Step 1: Detect the ATR-FTIR data and glycogen content of oyster tissue dry powder from different origins; Step 2: Import the ATR-FTIR data in step 1 into the spectrum analysis software, and perform second-order derivative preprocessing to obtain preprocessed spectrum data of the origin; The ATR-FTIR data in step 1 is imported into the spectral analysis software, and multivariate scattering correction and second-order derivative preprocessing are performed to obtain preprocessed spectral data of glycogen content; Step 3: Establishing a partial least squares discriminant analysis model of the origin based on the pre-processed spectral data and the origin, i.e., the origin discriminant model; A synergistic interval partial least squares regression model was established based on the pre-processed spectral data and glycogen content data, thus obtaining a quantitative model for glycogen content. Step 4: using the methods of step 1 and step 2 to obtain the pre-processed spectral data of the origin of the oyster tissue dry powder and the pre-processed spectral data of glycogen of the sample to be tested, and input them into the origin discrimination model and the glycogen content quantitative model respectively, and output the origin and glycogen content of the sample to be tested respectively; Among them, the origins of oyster tissues include Qinzhou City, Guangxi Zhuang Autonomous Region, Zhanjiang City, Guangdong Province, Dongshan County, Fujian Province, Cangnan County, Zhejiang Province, Qinhuangdao City, Hebei Province, the Yellow Sea waters of Dalian City, Liaoning Province, the Bohai Sea waters of Dalian City, Liaoning Province, and Rushan City, Shandong Province; The spectral range of ATR-FTIR is 650~4000cm -1 ; For the quantitative model of glycogen content, the spectrum was divided into 25 intervals, and the interval combination method was 4. The selected intervals were 2794-2928 cm -1 1320~1588cm -1 、1052~1186cm -1 ; The method for establishing the origin discrimination model specifically includes: The Kennard-Stone algorithm is used to divide the pre-processed spectral data and the origin of the oyster tissue dry powder from each origin into a training set and a validation set, wherein the training set is 3 / 4 of the total number of samples, the training set is used to establish a partial least squares discriminant analysis model of the origin, and the validation set is used to verify the partial least squares discriminant analysis model of the origin; The origin preprocessing data obtained in step 2 is used as an input variable, and the origin variable is used as an output variable to be imported into the statistical analysis software for normalization processing to obtain the origin discrimination model.

2. The method for simultaneously detecting the origin and glycogen content of oysters based on ATR-FTIR according to claim 1, characterized in that: The method for establishing the glycogen content quantitative model specifically includes: The pre-processed spectral data and glycogen content data of oyster tissue dry powder from various origins were divided into a training set and a validation set using the Kennard-Stone algorithm, wherein the training set was 3 / 4 of the total number of samples, the training set was used to establish a synergistic interval partial least squares regression model, and the validation set was used to validate the synergistic interval partial least squares regression model; The pre-processed spectral data of glycogen content is used as an input variable, and the glycogen content is used as an output variable to import into statistical analysis software to establish the glycogen content quantitative model.

3. The method for simultaneously detecting the origin and glycogen content of oysters based on ATR-FTIR according to claim 1, characterized in that: The method for preparing oyster tissue dry powder comprises: washing the oyster sample, dissecting the complete oyster tissue, freeze-drying, and grinding until the particle size of the oyster tissue is less than 100 meshes.

Images