Soybean oil producing area identification method based on multi-dimensional lipidomics characteristic fusion

Abstract

The invention relates to the technical field of food detection, particularly discloses a soybean oil production place identification method based on multi-dimensional lipidomics feature fusion, and aims at solving the problems that in the prior art, soybean oil production place identification is low in accuracy and depends on the content of a single component. The method comprises the following steps: collecting soybean oil samples from different producing areas and extracting grease; carrying out lipidomics analysis by adopting an ultra-high performance liquid chromatography-mass spectrometry technology; calculating key proportion indexes such as ratios of diglyceride to triglyceride, saturated and unsaturated fatty acids, unsaturated and saturated glyceride, phosphatidyl ethanolamine to total phospholipids and the like; screening features with significant statistical significance through orthogonal partial least square discriminant analysis; and constructing a production place identification model by using a random forest or a support vector machine algorithm. By adopting the technical scheme, high-precision identification of the soybean oil producing area can be realized, the identification accuracy of the soybean oil produced in China is as high as 98.72%, the method is obviously superior to the prior art, and the method is suitable for food traceability and market supervision.

Description

Technical Field

[0001] This invention belongs to the field of food quality testing and analysis technology, specifically a method for identifying the origin of soybean oil based on the fusion of multi-dimensional lipidomics characteristics. Background Technology

[0002] The field of food quality and safety testing involves two major issues: the risks posed by toxic and harmful substances and the authenticity of products. Regarding the detection of harmful substances in food, numerous domestic and international publications report standards and testing methods for detecting these substances. As for the authenticity of food quality, this issue has gradually attracted the attention and concern of consumers both domestically and internationally over the past decade, becoming a major focus and challenge in the field of food quality testing. Currently, technologies for detecting food authenticity mainly include fingerprinting techniques such as ultraviolet spectroscopy and infrared spectroscopy; atomic spectroscopy techniques such as atomic absorption, emission, and fluorescence; isotope mass spectrometry; high-resolution mass spectrometry; nuclear magnetic resonance (NMR); Raman spectroscopy; and omics technologies that emerged in the 1990s. Food omics within omics technologies encompasses epigenomics, transcriptomics, proteomics, metabolomics, and lipidomics. In the field of food inspection, proteomics, metabolomics, and lipidomics are the most commonly used omics technologies. These technologies can support the analysis of the authenticity of food's functional components, the content of nutritional components, and the traceability of its origin. In the field of authenticating the origin of food, isotope mass spectrometry and omics technology are considered to be two relatively reliable identification technologies. However, there are still few methods and patents for using omics technology to trace the origin of soybean oil.

[0003] Patent application CN111272861A discloses a MALDI-TOF method for detecting peptides in food. This method extracts peptides from a food matrix, eliminates interfering substances, and uses MALDI-TOF MS to detect the distribution of peptides in the food. The mass spectrometry is performed in positive ion mode with 337nm N2 laser irradiation at 50%-70% energy. Compared with traditional peptide detection methods, this method has simpler sample pretreatment, higher tolerance to sample impurities, higher sensitivity, better precision, wider detection range, shorter processing time, larger information content, and simpler analysis. Furthermore, this method can detect multiple samples simultaneously, possessing high throughput characteristics, providing a guarantee for rapid detection of nutritional assessment, origin traceability, adulteration, and quality process control in food. However, this method has difficulties in detecting peptides in some foods, especially the complex detection of peptides from different origins of the same product, making it difficult to directly apply to the traceability identification of soybean oil origin.

[0004] However, this method does not involve the construction of a lipidomics-based soybean oil origin traceability model. Summary of the Invention

[0005] The purpose of this invention is to provide a method for identifying the origin of soybean oil based on the fusion of multi-dimensional lipidomics features, which can effectively solve the problems in the background art mentioned above.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for identifying the origin of soybean oil based on the fusion of multi-dimensional lipidomics features includes the following specific steps: Step 1: Collect soybean oil samples from different origins, and perform lipidomics analysis on the soybean oil samples using ultra-high performance liquid chromatography coupled with high resolution mass spectrometry to obtain lipid data of soybean oil from different origins. The lipid data are then processed using mass spectrometry data analysis software. Step 2: Analyze the obtained lipid data using orthogonal partial least squares discriminant analysis (PLS) to construct an PLS model; select key lipid features with statistical significance based on the VIP value and p-value in the PLS model. Step 3: Calculate feature data values ​​based on key lipid data characteristics, the feature data values ​​including - First characteristic value: The ratio of total diglycerides to total triglycerides; -Second characteristic value: Total value of unsaturated glycerides; -Third characteristic value: Total phospholipid content; - Fourth characteristic value: Total value of saturated fatty acids. Step 4: Using feature data values, construct a soybean oil origin identification model through machine learning algorithms to achieve accurate identification of the origin of unknown soybean oil samples.

[0007] Preferably, the different origins in step 1 are selected from two or more of the following: the United States, Argentina, Uruguay, Brazil, and China.

[0008] Preferably, the ultra-high performance liquid chromatography (UHPLC) conditions used in step 1 specifically include: The chromatographic column is a C18 column with dimensions of 100 mm × 3 mm; Mobile phase A is a mixed solution of methanol, acetonitrile and water in a volume ratio of 1:1:1, containing 5 mM ammonium acetate; Mobile phase B is isopropanol, which contains 5 mM ammonium acetate; The gradient elution program was set as follows: 0-1 min, maintaining the proportion of mobile phase B at 20%; 1-5 min, linearly increasing the proportion of mobile phase B from 20% to 60%; 5-25 min, linearly increasing the proportion of mobile phase B from 60% to 98%; 25-27 min, maintaining the proportion of mobile phase B at 98%; 27-27.1 min, rapidly decreasing the proportion of mobile phase B from 98% to 20%; 27.1-30 min, maintaining the proportion of mobile phase B at 20% for column equilibration. The mobile phase flow rate was set to 0.3 ml / min; the injection plate temperature was controlled at 15℃; the column temperature was maintained at 50℃; and the injection volume was 1 μl.

[0009] Preferably, the mass spectrometry conditions for high-resolution mass spectrometry detection in step 1 specifically include: using an electrospray ionization source, and alternately acquiring data in positive ion mode and negative ion mode; setting the curtain gas pressure to 35 kPa; setting the nebulizer gas pressure to 55 kPa; setting the auxiliary heating gas pressure to 55 kPa; setting the ion source voltage to +5500 V in positive ion mode and -4500 V in negative ion mode; maintaining the ion source temperature at 550 °C; setting the declustering voltage to 80 V; setting the collision energy to 40 eV, with a collision energy fluctuation range of ±20 eV; setting the time-of-flight mass spectrometry mass scan range to 200-1250 m / z; setting the first-stage mass spectrometry accumulation time to 0.2 s; performing 30 product ion scans in each scan cycle; setting the accumulation time for each second-stage mass spectrometry scan to 20 ms; and setting the second-stage mass spectrometry scan range to 100-1250 m / z.

[0010] Preferably, in step 2, based on the variable importance projection values ​​of the orthogonal partial least squares discriminant analysis model and the significance values ​​of the analysis of variance, the proportion feature with a VIP value greater than 1 and a p value less than 0.05 is selected as the key identification indicator for distinguishing soybean oil from different origins.

[0011] Preferably, in step 3, the diglycerides include DG 51:8, DG 51:9, DG 51:7, DG 39:8, DG 41:10, DG 41:9, DG 38:1, DG 41:11, DG 30:0, DG 52:8, DG 45:3, DG 52:2, DG 36:0, DG 34:0, and DG 32:0; the triglycerides include TG 55:6, TG 48:2, TG 52:5;O, TG 52:6;O, TG 54:7;O, TG 54:8;O, TG 52:4;O, TG 54:6;O, TG 54:5;O, TG 54:4, TG 55:5, TG 53:5, and TG 55:4. In this invention, the ratio of diglycerides to triglycerides is used to reflect the compositional characteristics of glycerides. The ratio of the total diglyceride value to the total triglyceride value is obtained by calculating the ratio of the sum of the peak areas of all diglyceride components to the sum of the peak areas of all triglyceride components in each sample. This index can effectively characterize the differences in glyceride metabolic pathways of soybean oil from different origins.

[0012] Preferably, the unsaturated glycerides in step 3 include TG (18:3_18:2_19:2_19:1), TG (18:3_18:2_16:0;O), TG (18:2_18:2_19:1), TG (17:1_17:2_18:2_18:1), TG (17:2_18:2_18:2), TG (16:0_18:3_18:3;O), TG (18:2_18:2_18:3;O), DG (18:2_18:2), TG (18:3_18:3_18:2;O), TG (18:2_18:2_18:2;O), TG (18:1_18:2_18:2;O), TG (18:2_18:1_19:1), and DG (18:1_18:1). This invention obtains the total value of unsaturated glycerides by calculating the sum of the peak areas of all unsaturated glyceride components in each sample. This index can characterize the differences in desaturase activity during oil synthesis in soybeans from different origins.

[0013] Preferably, the saturated fatty acids in step 3 include FA 30:0, FA 32:0, FA 22:0, and FA 31:0.

[0014] Preferably, the phospholipids in step 3 include NAE 16:2 and NAE 14:0. In this invention, the total phospholipid content is obtained by calculating the sum of the peak areas of the phospholipid components in each sample. This indicator can reflect the specificity of soybeans from different producing areas in phospholipid metabolic pathways.

[0015] Preferably, in step S4, the machine learning classifier employs one or more combinations of Support Vector Machine (SVM), Random Forest (RF), or XGBoost algorithms, and the training process of the classifier includes the following steps: - Use the multidimensional feature vectors of standard soybean oil samples as the training set to label their corresponding place of origin; - Divide the training set data into a training subset and a validation subset in a 7:3 ratio; - Train the classifier using a training subset and adjust the classifier's hyperparameters using a validation subset; - After training, the classifier's performance is evaluated using a test set. Evaluation metrics include accuracy, recall, and F1 score.

[0016] Compared with the prior art, the present invention has the following beneficial effects: This invention utilizes the systematic mining of lipid category ratio characteristics to identify the origin of soybean oil, overcoming the limitations of existing technologies that only focus on the absolute content of a single lipid component, and significantly improving the identification accuracy. It leverages the significant differences in diglyceride and triglyceride content, as well as the distinct levels of unsaturated glycerides, phospholipids, and saturated fatty acids in soybean oil from different regions. These characteristics reflect regional differences more effectively than single lipid components, enabling the model to maintain high identification capability even when facing regions with similar climatic and geographical conditions. The constructed identification model achieves an accuracy of 98.72% for identifying soybean oil from China, far exceeding the approximately 85% of existing technologies, making it particularly suitable for verifying the authenticity of origin in market supervision and trade disputes. This method has good universality and can be extended to the identification of other vegetable oils, providing new ideas and methods for food traceability technology. Attached Figure Description

[0017] Figure 1 This is a total ion chromatogram of soybean oil from different origins detected in an embodiment of the present invention; Figure 2 This is a total ion chromatogram of soybean oil from different origins detected in an embodiment of the present invention; Figure 3 This is a PLA-DA diagram of differentially expressed proteins in soybean oil from different origins in this embodiment of the invention; Figure 4 This is a heatmap (HOT50) of cluster analysis of differential proteins in soybean oil from different origins in this embodiment of the invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments. Example

[0019] This embodiment provides a method for identifying the origin of soybean oil based on the fusion of multi-dimensional lipidomics features, including the following specific steps: Step 1: Collect soybean oil samples from different origins, and perform lipidomics analysis on the soybean oil samples using ultra-high performance liquid chromatography coupled with high resolution mass spectrometry to obtain lipid data of soybean oil from different origins. The lipid data are then processed using mass spectrometry data analysis software. Specifically, soybean oil samples from different origins were collected, and the oil was extracted using Soxhlet extraction to ensure consistent sample quality. In this example, the origins were selected from the United States, Argentina, Uruguay, and China, with 10 samples collected from each origin, totaling 40 soybean oil samples. Soybeans were ground into a fine powder (2g was accurately weighed) and wrapped in grease-free filter paper. The filter paper package was placed in the extraction tube of the Soxhlet extractor, and the extraction flask and condenser of known weight were connected, ensuring the apparatus was sealed. An appropriate amount of petroleum ether or anhydrous diethyl ether was added to the extraction flask, and the apparatus was placed in a water bath for heating. After the solvent boiled, the vapor rose, condensed, and dripped into the extraction tube to dissolve the oil. When the liquid level reached the height of the siphon tube, the oil-containing solvent flowed back into the extraction flask. This process was repeated for 6-8 hours until extraction was complete. After extraction, heating was stopped, and the extraction tube and condenser were removed after cooling. The solvent in the extraction flask was evaporated to dryness in a water bath in a fume hood. The test solution was diluted with a 1:1 mixture of methanol and isopropanol containing 5 mM ammonium acetate. All samples were stored at -20°C immediately after extraction for later use.

[0020] The chromatographic conditions for detection using ultra-high performance liquid chromatography coupled with high resolution mass spectrometry in this embodiment specifically include: The chromatographic column is a C18 column with dimensions of 100 mm × 3 mm; Mobile phase A is a mixed solution of methanol, acetonitrile and water in a volume ratio of 1:1:1, containing 5 mM ammonium acetate; Mobile phase B is isopropanol, which contains 5 mM ammonium acetate; The gradient elution program was set as follows: 0-1 min, maintaining the proportion of mobile phase B at 20%; 1-5 min, linearly increasing the proportion of mobile phase B from 20% to 60%; 5-25 min, linearly increasing the proportion of mobile phase B from 60% to 98%; 25-27 min, maintaining the proportion of mobile phase B at 98%; 27-27.1 min, rapidly decreasing the proportion of mobile phase B from 98% to 20%; 27.1-30 min, maintaining the proportion of mobile phase B at 20% for column equilibration. The mobile phase flow rate was set to 0.3 ml / min; the injection plate temperature was controlled at 15℃; the column temperature was maintained at 50℃; and the injection volume was 1 μl. This chromatographic condition design fully considers the polarity differences of various lipid molecules in soybean oil. A gradient elution program was used to achieve efficient separation from low-polarity triglycerides to high-polarity phospholipids, ensuring clear differentiation of various lipid components in the chromatogram. The TIC chromatograms of QC in both positive and negative ion modes showed excellent reproducibility. Figure 1 and Figure 2 .

[0021] The specific mass spectrometry conditions for ultra-high performance liquid chromatography coupled with high resolution mass spectrometry (UHPLC) in this embodiment are as follows: an electrospray ionization source is used, and data are acquired alternately in positive ion mode and negative ion mode; the curtain gas pressure is set to 35 kPa; the nebulizer gas pressure is set to 55 kPa; the auxiliary heating gas pressure is set to 55 kPa; the ion source voltage is +5500 V in positive ion mode and -4500 V in negative ion mode; the ion source temperature is maintained at 550 °C; the declustering voltage is set to 80 V; the collision energy is set to 40 eV, with a fluctuation range of ±20 eV; the mass scan range of time-of-flight mass spectrometry is set to 200-1250 m / z; the accumulation time of the first-stage mass spectrometry is 0.2 s; 30 product ion scans are performed in each scan cycle, and the accumulation time of each second-stage mass spectrometry scan is 20 ms, with the second-stage mass spectrometry scan range set to 100-1250 m / z. Mass spectrometry data are processed using specialized lipidomics analysis software for peak extraction, alignment, normalization, and lipid annotation, ultimately generating a data matrix containing peak area information for each lipid molecule, providing the raw data foundation for subsequent proportional feature calculations.

[0022] Step 2: Analyze the obtained lipid data using orthogonal partial least squares discriminant analysis (PLS) to construct an PLS model; select key lipid features with statistical significance based on the VIP value and p-value in the PLS model. Specifically, orthogonal partial least squares discriminant analysis (PLS-DA) was used to analyze the obtained lipid data. An PLS-DA model was constructed, and key lipid data with statistical significance were selected based on the variable importance projection values ​​(VIP values) and the significance values ​​(p-values) of the analysis of variance. In this embodiment, the proportion of lipids with VIP values ​​greater than 1 and p-values ​​less than 0.05 were selected as key identification indicators for differentiating soybean oils from different origins. The characteristic lipids are shown in Table 1 below.

[0023] Table 1. Significant Differences in Lipid Scale The PLS-DA result diagram in this embodiment is as follows: Figure 3As shown in the figure, AGT represents Argentina, GC represents China, MG represents the United States, and WLG represents Uruguay. It can be seen from the figure that there are differences in the four groups of soybean oil components. Figure 4 This is a heatmap (top 50) of lipid clustering analysis in this embodiment. It can be seen from the heatmap that the differential components can significantly distinguish the four groups of soybean oil.

[0024] Step 3: Calculate feature data values ​​based on key lipid data characteristics, the feature data values ​​including - First characteristic value: The ratio of total diglycerides to total triglycerides; -Second characteristic value: Total value of unsaturated glycerides; -Third characteristic value: Total phospholipid content; - Fourth characteristic value: Total value of saturated fatty acids.

[0025] Specifically, in this embodiment, diglycerides include DG 51:8, DG 51:9, DG 51:7, DG39:8, DG 41:10, DG 41:9, DG 38:1, DG41:11, DG 30:0, DG 52:8, DG 45:3, DG52:2, DG 36:0, DG 34:0, and DG 32:0. Triglycerides include TG55:6, TG 48:2, TG 52:5; 0, TG52:6; 0, TG 54:7; 0, TG54:8; 0, TG 52:4; 0, TG54:6; 0, TG 54:5; 0, TG 54:4, TG 55:5, TG 53:5, and TG 55:4. In this invention, the ratio of diglycerides to triglycerides is used to reflect the compositional characteristics of glycerides. The ratio of the total diglyceride value to the total triglyceride value is obtained by calculating the ratio of the sum of the peak areas of all diglyceride components to the sum of the peak areas of all triglyceride components in each sample. This index can effectively characterize the differences in glyceride metabolic pathways of soybean oil from different origins.

[0026] Unsaturated glycerides include TG (18:3_18:2_19:2_19:1), TG (18:3_18:2_16:0;O), TG (18:2_18:2_19:1), TG (17:1_17:2_18:2_18:1), TG (17:2_18:2_18:2), TG (16:0_18:3_18:3;O), TG (18:2_18:2_18:3;O), DG (18:2_18:2), TG (18:3_18:3_18:2;O), TG (18:2_18:2_18:2;O), TG (18:1_18:2_18:2;O), TG (18:2_18:1_19:1), DG (18:1_18:1). This invention obtains the total value of unsaturated glycerides by calculating the sum of the peak areas of all unsaturated glyceride components in each sample. This index can characterize the differences in desaturase activity during oil synthesis in soybeans from different producing areas.

[0027] Saturated fatty acids include FA 30:0, FA 32:0, FA 22:0, and FA 31:0. Phospholipids include NAE 16:2 and NAE 14:0. In this invention, the total phospholipid content is obtained by calculating the sum of the peak areas of the phospholipid components in each sample. This indicator can reflect the specificity of soybeans from different producing areas in the phospholipid metabolic pathway.

[0028] The characteristic data values ​​of each production area in this embodiment are shown in Table 2-5 below: Table 2 Characteristic data values ​​of soybean oil samples from China Table 3 Characteristic data values ​​of soybean oil samples from the United States Table 4. Characteristic data values ​​of soybean oil samples from Argentina Table 5. Characteristic data values ​​of soybean oil samples from Uruguay It should be noted that the content of unsaturated glycerides, phospholipids and saturated fatty acids in the characteristic data values ​​have been processed to make their orders of magnitude within the range described in the table, and do not represent their actual content.

[0029] Step 4: Using feature data values, construct a soybean oil origin identification model through machine learning algorithms to achieve accurate identification of the origin of unknown soybean oil samples.

[0030] Specifically, using the selected feature data values, a soybean oil origin identification model was constructed through machine learning algorithms to accurately identify the origin of unknown soybean oil samples. The machine learning algorithms were implemented using the R programming language. The Random Forest algorithm and the Support Vector Machine algorithm were analyzed using the Support Vector Machine extension package. All soybean oil samples were randomly split into training and test sets, with the training set comprising 70% of the data matrix and the test set comprising 30%, used for model construction, parameter tuning, and performance evaluation. Simultaneously, an additional 20 soybean oil samples from unknown origins were collected for blind testing through external sample validation. The overall accuracies of the Random Forest and Support Vector Machine models were 93.5% and 96.0%, respectively, with an identification accuracy of 98.0% for soybean oil samples from China, verifying the reliability of this method in practical applications.

[0031] While several embodiments of the present invention have been provided herein, those skilled in the art should understand that modifications can be made to the embodiments without departing from the spirit of the invention. The above embodiments are merely exemplary and should not be construed as limiting the scope of the invention.

Claims

1. A method for identifying the origin of soybean oil based on the fusion of multi-dimensional lipidomics features, characterized in that, The method includes the following steps: Step 1: Collect soybean oil samples from different origins, and perform lipidomics analysis on the soybean oil samples using ultra-high performance liquid chromatography coupled with high resolution mass spectrometry to obtain lipid data of soybean oil from different origins. The lipid data are then processed using mass spectrometry data analysis software. Step 2: Analyze the obtained lipid data using orthogonal partial least squares discriminant analysis (PLS) to construct an PLS model; select key lipid features with statistical significance based on the VIP value and p-value in the PLS model. Step 3: Calculate feature data values ​​based on key lipid data characteristics, the feature data values ​​including - First characteristic value: The ratio of total diglycerides to total triglycerides; -Second characteristic value: Total value of unsaturated glycerides; -Third characteristic value: Total phospholipid content; - Fourth characteristic value: Total saturated fatty acid content; Step 4: Using feature data values, construct a soybean oil origin identification model through machine learning algorithms to achieve accurate identification of the origin of unknown soybean oil samples.

2. The method for identifying the origin of soybean oil based on multi-dimensional lipidomics feature fusion as described in claim 1, characterized in that, The different origins in step 1 are selected from two or more of the following: the United States, Argentina, Uruguay, Brazil, and China.

3. The method for identifying the origin of soybean oil based on multi-dimensional lipidomics feature fusion as described in claim 2, characterized in that, The chromatographic conditions for ultra-high performance liquid chromatography in step 1 specifically include: The chromatographic column is a C18 column with dimensions of 100 mm × 3 mm; Mobile phase A is a mixed solution of methanol, acetonitrile and water in a volume ratio of 1:1:1, containing 5 mM ammonium acetate; Mobile phase B is isopropanol, which contains 5 mM ammonium acetate; The gradient elution program was set as follows: 0-1 min, maintaining the proportion of mobile phase B at 20%; 1-5 min, linearly increasing the proportion of mobile phase B from 20% to 60%; 5-25 min, linearly increasing the proportion of mobile phase B from 60% to 98%; 25-27 min, maintaining the proportion of mobile phase B at 98%; 27-27.1 min, rapidly decreasing the proportion of mobile phase B from 98% to 20%; 27.1-30 min, maintaining the proportion of mobile phase B at 20% for column equilibration. The mobile phase flow rate was set to 0.3 ml / min; the injection plate temperature was controlled at 15℃; the column temperature was maintained at 50℃; and the injection volume was 1 μl.

4. The method for identifying the origin of soybean oil based on multi-dimensional lipidomics feature fusion as described in claim 3, characterized in that, The specific mass spectrometry conditions for high-resolution mass spectrometry in step 1 include: using an electrospray ionization source, and alternately acquiring data in positive ion mode and negative ion mode; setting the curtain gas pressure to 35 kPa; setting the nebulizer gas pressure to 55 kPa; setting the auxiliary heating gas pressure to 55 kPa; setting the ion source voltage to +5500 V in positive ion mode and -4500 V in negative ion mode; maintaining the ion source temperature at 550 °C; setting the declustering voltage to 80 V; setting the collision energy to 40 eV, with a fluctuation range of ±20 eV; setting the mass scan range of time-of-flight mass spectrometry to 200-1250 m / z; setting the first-stage mass spectrometry accumulation time to 0.2 s; performing 30 product ion scans in each scan cycle; setting the accumulation time of each second-stage mass spectrometry scan to 20 ms; and setting the second-stage mass spectrometry scan range to 100-1250 m / z.

5. The method for identifying the origin of soybean oil based on multi-dimensional lipidomics feature fusion as described in claim 1, characterized in that, In step 2, based on the variable importance projection values ​​of the orthogonal partial least squares discriminant analysis model and the significance values ​​of the analysis of variance, the proportion of features with a VIP value greater than 1 and a p value less than 0.05 is selected as the key identification indicator for distinguishing soybean oil from different origins.

6. The method for identifying the origin of soybean oil based on multi-dimensional lipidomics feature fusion as described in claim 1, characterized in that, In step 3, the diglycerides include DG 51:8, DG 51:9, DG 51:7, DG 39:8, DG 41:10, DG 41:9, DG 38:1, DG 41:11, DG 30:0, DG 52:8, DG 45:3, DG 52:2, DG 36:0, DG 34:0, and DG 32:0; the triglycerides include TG 55:6, TG 48:2, TG 52:5;O, TG 52:6;O, TG 54:7;O, TG 54:8;O, TG 52:4;O, TG 54:6;O, TG 54:5;O, TG 54:4, TG 55:5, TG 53:5, and TG 55:

4.

7. The method for identifying the origin of soybean oil based on multi-dimensional lipidomics feature fusion as described in claim 6, characterized in that, The unsaturated glycerides in step 3 include TG (18:3_18:2_19:2_19:1), TG (18:3_18:2_16:0;O), TG (18:2_18:2_19:1), TG (17:1_17:2_18:2_18:1), TG (17:2_18:2_18:2), TG (16:0_18:3_18:3;O), TG (18:2_18:2_18:3;O), DG (18:2_18:2), TG (18:3_18:3_18:2;O), TG (18:2_18:2_18:2;O), TG (18:1_18:2_18:2;O), TG (18:2_18:1_19:1), and DG (18:1_18:1).

8. The method for identifying the origin of soybean oil based on multi-dimensional lipidomics feature fusion as described in claim 7, characterized in that, The saturated fatty acids in step 3 include FA 30:0, FA 32:0, FA 22:0, and FA 31:

0.

9. The method for identifying the origin of soybean oil based on multi-dimensional lipidomics feature fusion as described in claim 8, characterized in that, The phospholipids in step 3 include NAE 16:2 and NAE 14:0.

Images