A method for identifying animal-derived components in meat products based on molecular spectroscopy

Abstract

A kind of animal-derived component discrimination analysis method in meat product based on molecular spectrum relates to chemical analysis technical field.The present application uses the fat basic spectrum obtained from different parts of the same species fat, after spectrum pretreatment, establishes the basic library of the fat of this species;Using the same spectrum measurement condition, the spectrum is collected, using the same spectrum pretreatment method, the spectrum is processed, and the basic library of the fat of other species is established;Using the same spectrum measurement condition, the spectrum of the fat of the sample to be measured is collected and using the same spectrum pretreatment method, the spectrum of the fat of the sample to be measured is processed, the similarity of the spectrum of the sample to be measured and the spectrum of each species basic library is calculated, and the animal-derived species is determined according to the similarity.The present application eliminates the interference of fluorescence signal and the signal overlap of different fat spectrum with similar structure on species identification.Compared with modeling combined with molecular spectrum and chemometrics method, modeling is not needed, the number of samples used is less, the operation is simple, and the rapid and accurate discrimination of animal-derived components in meat product is realized.

Description

Technical Field

[0001] This invention relates to the field of chemical analysis, and more specifically, to the design of an analytical method for identifying animal-derived components in meat products based on molecular spectroscopy. Background Technology

[0002] Meat is rich in protein and provides all the essential amino acids, fatty acids, vitamins, and trace elements the human body needs. Therefore, compared to grains, fruits, and vegetables, meat is generally considered to have a higher nutritional value. With rising living standards, meat products have become a necessity on our tables. However, with the increasing demand for meat products, some producers and distributors may engage in adulteration to gain economic benefits. Furthermore, there are individuals with strict requirements regarding animal meat products due to cultural beliefs and vegetarianism. Therefore, the safe assessment of meat quality is extremely important.

[0003] Currently, commonly used methods for testing meat products include sensory evaluation, PCR, and liquid chromatography / mass spectrometry (GB / T22210-2008 "Sensory Evaluation Standard for Meat and Meat Products", NY / T 3904-2021 "Detection of Heterocyclic Amines in Meat and Meat Products by Liquid Chromatography-Tandem Mass Spectrometry", GB / T23815-2000 "Qualitative PCR Detection Method for Plant Components in Pork Products"). Sensory evaluation relies on the human senses, including the eyes, nose, mouth (including lips and tongue), and hands, to assess food. Visual inspection (eyes) is used to examine the appearance, muscle color, and fat color of meat products; tactile inspection (hands) is used to examine the muscle elasticity of meat products; olfactory inspection (nose) is used to identify the odor of meat products; and taste inspection (mouth and tongue) is used to identify processed meat products. Sensory evaluation results are greatly affected by human factors and external environmental interference, impacting the objectivity of the results. PCR (Polymerase Chain Reaction) is a DNA-based detection method. DNA, or deoxyribonucleic acid, is a fundamental component of animal cell chromosomes, encoding genetic information in proteins and RNA molecules that determine species traits. PCR refers to the process of amplifying complementary daughter DNA fragments to the parental template DNA fragment in vitro under the catalysis of DNA polymerase. A key factor in using PCR for meat adulteration identification is selecting appropriate species-specific target genes. PCR amplification technology has high detection sensitivity, capable of detecting adulteration as low as 0.001% in meat. However, PCR technology is susceptible to DNA degradation and complex mechanisms, and is only suitable for laboratory testing. Detection and analysis require a long time, demanding strict requirements on instruments and operators. Furthermore, if the PCR primer design is inappropriate, the incidence of false positives and false negatives after specific amplification is relatively high. Liquid chromatography / mass spectrometry (LC / MS) is a biomolecular mass spectrometry technique based on proteomics. It identifies meat products by detecting specific proteins from different animal sources. Compared with PCR, biomolecular mass spectrometry offers faster, more accurate, and more specific detection. However, biomolecular mass spectrometry instruments are expensive, hindering its widespread application. Therefore, there is an urgent need for an analytical method that is fast, accurate, inexpensive, and easy to operate for identifying animal-derived components in meat products.

[0004] Molecular spectroscopy can reflect the composition and structure of meat products at the molecular level. With the development of instrument technology, it is possible to conveniently and quickly acquire sample spectra. Currently, molecular spectroscopy combined with traditional chemometrics methods is often used in qualitative research on meat products. Commonly used traditional chemometrics methods include PLS-DA, SVM, and SIMCA. PLS-DA (Partial Least Squares Discriminant Analysis) is a supervised pattern recognition method based on PLS (Partial Least Squares). The analysis results are intuitive and clear, with high prediction accuracy. However, it is not suitable for situations with strong linear correlation and many nonlinear interference factors. The acquisition of molecular spectra of meat products is greatly affected by measurement conditions, sample morphology, and freezing time. With too many interference factors, PLS-DA is prone to misjudgment. At the same time, the number of fat samples collected for modeling is very large, ranging from hundreds to thousands. Even with such a large sample size, it is difficult to cover all the spectra required for modeling, resulting in a large workload and making it difficult to apply in practice. SVM (Support Vector Machine) has good adaptability to high-dimensional processing space and can adapt to the multidimensional computation problems of molecular spectroscopy. However, SVM is usually used for binary classification and cannot be directly used for multi-class classification. In practical applications, it is necessary to classify samples from multiple species. The application of SVM in practical scenarios is limited. At the same time, SVM often suffers from overfitting, which affects the judgment of results.

[0005] Therefore, developing a method to address the bottleneck issues encountered in the rapid identification and analysis of animal-derived species in meat products using molecular spectroscopy is of significant practical importance. This patent discloses a novel qualitative method that establishes a basic lipid spectral library for a specific species by preprocessing lipid spectra obtained from different parts of the same species. Similarly, lipid spectra from other species are acquired using the same spectral measurement conditions and preprocessing methods. The similarity between the measured spectrum and the basic spectra of each species is calculated, and the animal-derived species is determined based on this similarity. This invention establishes basic lipid libraries for different species and uses similarity to identify the animal-derived species in meat products, eliminating interference from overlapping fluorescence signals and structurally similar lipid spectral signals. It overcomes the drawbacks of traditional methods, such as high sample preparation requirements and long analysis times, and solves the problem of difficult modeling when combining molecular spectroscopy with traditional chemometrics, achieving rapid and accurate identification of animal-derived components in meat products. Summary of the Invention

[0006] The purpose of this invention is to provide an analytical method for identifying animal-derived components in meat products based on molecular spectroscopy. Under the same spectral measurement conditions, the collected fat spectra are compared with the similarity of fat baseline libraries for various species to identify animal-derived components in meat products. This method effectively solves the problems encountered by traditional methods in identifying animal-derived components in meat products, achieving rapid and accurate identification, and at least partially overcoming the shortcomings of existing methods.

[0007] The present invention provides an analytical method for identifying animal-derived components in meat products based on molecular spectroscopy, comprising the following steps:

[0008] A basic lipid spectral library for the species was established by preprocessing lipid spectra obtained from different parts of the same species.

[0009] Spectra were acquired using the same spectral measurement conditions and processed using the same preprocessing methods to establish a basic lipid database for other species.

[0010] The lipid spectra of the test samples were collected using the same spectral measurement conditions and preprocessed using the same methods. The animal origin of the test sample was determined by calculating the similarity between the spectrum of the test sample and the spectrum of each species' basic library.

[0011] Preferably, the spectrum includes at least one of infrared spectrum, near-infrared spectrum and Raman spectrum.

[0012] Preferably, the fat spectrum includes at least one band of the full spectrum of the verification sample.

[0013] Preferably, the spectral acquisition conditions of the sample to be tested should be consistent with those of the basic library.

[0014] Preferably, the similarity is expressed in one of the following forms: spectral residual, similarity coefficient, included angle, and trigonometric function value, or a combination of two or more.

[0015] Preferably, the spectral preprocessing method includes one of derivative, normalization, baseline correction, smoothing, etc., or a combination of two or more.

[0016] This invention relates to the field of chemical analysis technology and provides an analytical method for identifying animal-derived components in meat products based on molecular spectroscopy. The method includes: establishing a basic lipid spectral library for a given species by preprocessing lipid spectra obtained from different parts of the same species; establishing lipid spectral libraries for other species under the same spectral acquisition conditions and preprocessing methods; acquiring and preprocessing lipid spectra of the test sample under the same conditions; and determining the animal-derived species by calculating the similarity between the test sample's spectrum and the basic spectra of each species' library. This invention establishes basic lipid libraries for different species and identifies the animal-derived species in meat products based on similarity, eliminating interference from overlapping fluorescence signals and structurally similar lipid spectral signals. Compared to molecular spectroscopy combined with chemometrics for modeling, this method eliminates the need for modeling, requires fewer samples, is simpler to operate, and achieves rapid and accurate identification of animal-derived components in meat products. Attached Figure Description

[0017] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0018] Figure 1 This is a flowchart of the analytical method for identifying animal-derived components in meat products based on molecular spectroscopy according to the present invention.

[0019] Figure 2 Repeatability and standard deviation of 5-point spectra collected from the same sample

[0020] Figure 3 A species base library composed of five types of meat.

[0021] Figure 4 To verify the comparison between the sample prediction results and the actual results Detailed Implementation

[0022] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. For ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0023] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0024] The meat products in this application embodiment can be pork, beef, mutton, chicken, duck, or horse meat, venison, etc. It is sufficient to obtain the basic fat library of the species by collecting fat spectra, which will not be elaborated here.

[0025] Traditional detection methods have high requirements for sample preparation and require a long time for analysis, making them unsuitable for rapid meat identification. Molecular spectroscopy for the identification of animal-derived components usually combines chemometric methods such as PLS-DV and SVM to build models. These models are complex to operate in terms of data modeling, require a large amount of data support, and need to continuously adjust the model parameters. The model building is quite complex and difficult, which affects the accuracy of the judgment.

[0026] Therefore, this application determines the animal origin species based on similarity. This method establishes a basic fat database for different species, calculates the similarity between the spectrum of the sample to be tested and the spectra of the fat databases of each species, and determines the animal origin species based on this similarity. No modeling is required, achieving rapid and accurate identification of animal origin components in meat products.

[0027] The species fat base library in this embodiment mainly comprises a set of fat spectral vectors from different parts of the same species. By using the aforementioned species fat base library and the fat spectrum of the sample to be tested, the similarity between the spectrum of the sample to be tested and the spectra of each species base library can be calculated, thereby determining the species of the unknown sample.

[0028] like Figure 1 As shown, this application provides an analytical method for identifying animal-derived components in meat products based on molecular spectroscopy, including the following steps.

[0029] S101: Basic lipid spectra obtained from different parts of the same species, and a basic lipid library for the species was established after preprocessing of the lipid spectra.

[0030] S102: Use the same spectral measurement conditions to acquire spectra and the same preprocessing methods to process spectra, and establish a basic fat library for other species;

[0031] S103: The lipid spectra of the test samples were collected using the same spectral measurement conditions and preprocessed using the same preprocessing methods. The animal origin of the test sample was determined by calculating the similarity between the spectrum of the test sample and the spectrum of each species' basic library.

[0032] In processing S101, since the composition of fat varies in different parts of the same species, the fat spectra of different parts of the same species will also differ. These differences in fat spectra provide the data foundation for the subsequent algorithm. In this embodiment, at least five parts of the sample to be tested are selected. Pork parts are classified according to "GB / T 9959.3-2019 Fresh and Frozen Pork and Pork By-products Part 3: Pork Sliced ​​by Part", beef parts are classified according to "GB / T 17238-2022 Fresh and Frozen Sliced ​​Beef", mutton parts are classified according to "NY / T 1564-2021 Technical Specification for Livestock and Poultry Meat Slicing (Mutton)", duck parts are classified according to "NY / T 3962-2021 Technical Specification for Livestock and Poultry Meat Slicing (Duck)", and chicken parts are classified according to "GB / T 24864 Chicken Carcass Slicing". Fat spectra of different parts of the same species are collected, and after preprocessing, a basic fat library for that species is established.

[0033] In processing S102, the same measurement conditions refer to the same laser wavelength, laser power, integration time, number of internal scans, and number of sample scans when collecting lipid databases for all species. The same preprocessing method refers to the same preprocessing method and spectral band selection used when processing lipid databases for all species.

[0034] In processing S103, "same measurement conditions" means that the laser wavelength, laser power, integration time, number of internal scans, and number of sample scans used when collecting the lipid spectra of the sample to be tested are consistent with those used when collecting the basic lipid database of the species. "Same preprocessing methods" means that the preprocessing methods and spectral band selection used when processing the lipid spectra of the sample to be tested are consistent with those used when processing the basic lipid database of the species. The validated lipid spectra must be representative; this representativeness means that after preprocessing, the standard deviation between the five measured spectra from the same validated sample lipid sample must be less than one percent. The animal origin of the sample is determined by calculating the similarity between the spectrum of the sample to be tested and the basic databases of each species.

[0035] In the above process, the spectrum used includes at least one of infrared spectrum, near-infrared spectrum and Raman spectrum. Preferably, the spectrum is Raman spectrum because Raman spectrum has a fast testing speed, low cost, and little influence of moisture on the measurement results. Its resolution can also meet the requirements of fat spectrum of meat products.

[0036] As a preferred approach, when constructing a basic library of meat product fat spectra, all parts of the same species can be selected. This way, the basic library of meat product fat spectra can cover most of the fat spectra as much as possible, thereby providing detection data support for cases with diverse samples.

[0037] Preferably, the spectra in the basic library of meat product fat spectra and the fat spectra collected from the validation samples should both be representative, that is, five fat spectra should be collected from the same fat sample, and the five fat spectra should be in the 300cm band. -1 -3000cm -1 The standard deviation should be less than one percent.

[0038] Preferably, the similarity expression form between the fat spectrum of the sample to be tested and each fat base library is selected by choosing the angle value expression form. The fat base library is a hyperplane, and the fat spectrum of the sample to be tested is a vector. Calculating the angle value between the vector and the hyperplane can accurately reflect the similarity between the fat spectrum of the sample to be tested and the base library.

[0039] Preferably, when preprocessing the fat spectrum, a baseline correction combined with maximum value normalization is used to eliminate interference from fluorescence signals for species identification. The baseline correction utilizes the AIRPLS algorithm, adjusting the baseline correction level based on the fat spectrum by changing the values ​​of the smoothness parameter lambda, the fitting parameter order, the weight parameter wep, the asymmetry parameter p, and the maximum number of iterations itermax. This results in a set of parameters suitable for baseline correction of the fat spectrum: lambda = 10e5, order = 2, wep = 0.5, p = 0.01, itermax = 20.

[0040] Preferably, when performing spectral analysis on the sample to be tested, the obtained spectrum includes at least one band of the full spectrum. Since the content of saturated and unsaturated fatty acids varies among different species, resulting in different shapes and intensities of representative peaks, in a preferred embodiment, the representative peak positions of saturated and unsaturated fatty acids are concentrated in the band 670-1800 cm⁻¹ for the testing of the meat product fat baseline library and validation samples. -1 and 2700-3000cm -1 It can eliminate the overlap of spectral signals from different fats with similar structures and the interference of instrument noise on species identification.

[0041] Example 1

[0042] Sample preparation

[0043] Fat samples from different parts of meat products from five different species—pork, beef, lamb, chicken, and duck—were collected and used to construct different basic libraries of fat spectra.

[0044] Five pork fat samples, five beef fat samples, five mutton fat samples, five chicken fat samples, and five duck fat samples were randomly selected from the collected fat samples as validation samples.

[0045] Instruments and equipment

[0046] Portman 785 Raman Analyzer (manufactured by Xipaite (Beijing) Technology Co., Ltd.)

[0047] Spectral Acquisition

[0048] Using a Portman 785 Raman analyzer, the sample was thawed at room temperature for 30 minutes, and the Raman spectrum was acquired. Instrument warm-up time: 30 minutes. Raman analyzer measurement range: 200-3000 cm⁻¹ -1 The internal scan was performed 3 times, the integration time was 10000ms, the laser wavelength was 785nm, the laser power was 500mW, and the average value of the 3 single-point scans was taken. The sample was placed on the detection window, ensuring close contact between the sample and the window, and the spectrum acquisition was initiated by clicking "Acquisition Spectrum." Five sample points were collected from each sample. After preprocessing, the results are as follows: Figure 2 As shown in the figure, the repeatability of the samples is good, within the range of 300-300 cm. -1 The standard deviations are all less than one percent.

[0049] Raman spectra of fat from different parts of five different species were obtained using the above measurement conditions. Baseline correction and normalization were used to preprocess the Raman spectra of the fat from the five species, and the 670-1800 cm⁻¹ range was selected. -1 and 2700-3000cm -1 Bands constitute the basic fat reservoirs of different species, such as Figure 3 As shown.

[0050] Raman spectra of fat samples from 25 verification samples were collected under the same measurement conditions. The same band selection and preprocessing method as the basic library was used on the Raman spectra of the verification sample fat. The similarity between the spectrum of the sample to be tested and the spectrum of each species in the basic library was calculated using the method proposed in this patent. That is, the angle value between the verification sample and each basic library was calculated. The species corresponding to the smallest angle value is the species of the verification sample. The animal species is determined based on the angle value. Figure 4 To verify that the species determined by calculating similarity of the samples were all correct, it is shown that the method for identifying animal-derived components in meat products based on molecular spectroscopy provided by this invention can quickly and accurately determine the species of unknown samples.

Claims

1. An analytical method for identifying animal-derived components in meat products based on molecular spectroscopy, characterized in that: Molecular spectra of fat from different parts of the same species were obtained, and after preprocessing, a basic molecular spectral library of fat for that species was constructed. Molecular spectra of fat from different parts of other species were collected using the same spectral acquisition conditions, and the spectra were processed using the same preprocessing method to construct basic molecular spectral libraries of fat for each species. Molecular spectra of the test samples were collected using the same spectral acquisition conditions and processed using the same preprocessing method, and the similarity between the spectra and the basic molecular spectra of fat for each species was calculated. Based on the similarity values, the animal origin of the sample was determined. The same spectral acquisition conditions refer to the consistent laser wavelength, laser power, integration time, number of internal scans, and number of sample scans when acquiring the basic lipid libraries of all species. The same preprocessing method refers to the consistent preprocessing method and spectral band selection used when processing the basic lipid libraries of all species. The basic molecular spectral library includes the molecular spectra of saturated and unsaturated fatty acids with different carbon numbers. Saturated fatty acids with different carbon numbers include: C14:0, C18:0, and C20:

0. Unsaturated fatty acids with different carbon numbers include: C20:3n6, C22:1n9, C20:3n3, C18:1n9c, C18:2n6c.

2. The analytical method for identifying animal-derived components in meat products based on molecular spectroscopy as described in claim 1, characterized in that, The molecular spectra include at least one of infrared spectroscopy, near-infrared spectroscopy, and Raman spectroscopy.

3. The analytical method for identifying animal-derived components in meat products based on molecular spectroscopy as described in claim 1, characterized in that, The molecular spectrum includes at least one band of the full spectrum of the sample to be tested.

4. The analytical method for identifying animal-derived components in meat products based on molecular spectroscopy as described in claim 1, characterized in that, The similarity can be expressed in one of the following forms: spectral residual, similarity coefficient, included angle and its trigonometric function value, or a combination of two or more.

5. The analytical method for identifying animal-derived components in meat products based on molecular spectroscopy as described in claim 1, characterized in that, A variety of spectral preprocessing methods were used, including derivative, baseline correction, normalization, and smoothing, or a combination of two or more of them.

Images