A quantitative analysis method for potato amino acids using liquid chromatography-tandem mass spectrometry

Abstract

The present invention discloses a quantitative analysis method for potato amino acids using liquid chromatography-tandem mass spectrometry, which relates to the technical field of analysis and detection. The method comprises the following steps: collecting amino acid samples from potatoes and chemical property data thereof, preprocessing the acquired data, obtaining specific rotations of amino acid sample solutions from the potatoes based on the preprocessed amino acid samples from the potatoes and the chemical property data thereof in combination with a linear regression algorithm, classifying the amino acid sample solutions into types using historical amino acid sample data from the potatoes, obtaining liquid chromatography column distribution results based on the types of the amino acid sample solutions from the potatoes, and constructing a liquid chromatography column distribution model using the liquid chromatography column distribution results. Technologies such as the linear regression algorithm, convolutional neural network algorithm, and neural network algorithm in the method of the present invention are closely integrated with modern information technology, significantly enhancing the intelligence level of the quantitative analysis method for potato amino acids using liquid chromatography-tandem mass spectrometry.

Description

Technical Field

[0001] The present invention relates to the technical field of data fusion and information processing, in particular to a quantitative analysis method of potato amino acids by liquid chromatography-tandem mass spectrometry. Background Art

[0002] In the field of quantitative analysis of potato amino acids by liquid chromatography-tandem mass spectrometry, accurate analysis of potato amino acids is crucial. Traditional amino acid analysis methods have limitations such as low sensitivity, poor resolution, and the inability to simultaneously determine multiple amino acids, making it difficult to meet the needs of in-depth research on potato amino acids. Although liquid chromatography can achieve amino acid separation, its qualitative ability is limited when used alone. Mass spectrometry, with its high sensitivity, can accurately determine the molecular weight and structural information of compounds. Combining liquid chromatography and mass spectrometry in tandem, and integrating the advantages of both, provides technical support for potato amino acid research. Therefore, a quantitative analysis method for potato amino acids by liquid chromatography-tandem mass spectrometry has emerged.

[0003] Although the existing technology has made great progress in the quantitative analysis of potato amino acids by liquid chromatography-tandem mass spectrometry, there are still some problems that need to be optimized. During the quantitative analysis of potato amino acids by liquid chromatography-tandem mass spectrometry, potatoes contain a variety of amino acids with similar chemical properties, which have similar retention behaviors on liquid chromatography columns, resulting in overlapping chromatographic peaks, affecting the quantitative analysis results of potato amino acids by liquid chromatography-tandem mass spectrometry. Summary of the Invention

[0004] In view of the deficiencies in the prior art, the present invention provides a quantitative analysis method for potato amino acids by liquid chromatography-tandem mass spectrometry, which solves the problems in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented by the following technical solution: a quantitative analysis method of potato amino acids by liquid chromatography tandem mass spectrometry, comprising the following steps:

[0006] Step 1: Obtain amino acid samples and their chemical property data from potatoes, and pre-process the acquired data to provide data support for subsequent steps;

[0007] Step 2: Based on the obtained optical rotation angle of the amino acid sample solution in potatoes, combined with a linear regression algorithm, the specific optical rotation of the amino acid sample solution in potatoes is obtained;

[0008] Step 3: Determine the classification criteria for amino acid sample solutions based on historical data of amino acid samples in potatoes, providing data support for constructing a classification model for amino acid sample solutions;

[0009] Step 4: Based on the classification criteria of the amino acid sample solutions, combined with the convolutional neural network algorithm, a classification model of the amino acid sample solutions is constructed to obtain the types of the amino acid sample solutions in potatoes;

[0010] Step 5: Obtaining liquid chromatography column partitioning results based on the types of amino acid sample solutions in potatoes, and constructing a liquid chromatography column partitioning model using the liquid chromatography column partitioning results. This solves the problem that potatoes contain multiple amino acids with similar chemical properties, resulting in overlapping chromatographic peaks and affecting the quantitative analysis results of potato amino acids by liquid chromatography-tandem mass spectrometry;

[0011] Step 6: Combine the output results of the liquid chromatography column distribution model to perform quantitative analysis of potato amino acids by liquid chromatography tandem mass spectrometry.

[0012] A further improvement of the technical solution of the present invention is that in step 1, the process of obtaining the amino acid sample and the chemical property data of the amino acid sample includes:

[0013] Deploy different collection equipment to obtain amino acid samples and chemical property data from potatoes, wherein the collection equipment includes filter paper, centrifuge tubes, ultrasonic extractors, solid phase extraction columns, nitrogen blowdown devices, sample tubes, high performance liquid chromatographs, pH meters, and polarimeters;

[0014] The chemical property data of the amino acid sample include water delivery parameters, pH value and optical rotation angle of the amino acid sample solution in potatoes;

[0015] Select potatoes, rinse the potatoes with deionized water, dry the potatoes with filter paper, peel the potatoes and cut them into small pieces, freeze them in liquid nitrogen to obtain frozen potatoes, weigh the frozen potato samples, place the frozen potato samples in a centrifuge tube filled with an extraction solvent, place the centrifuge tube in an ultrasonic extractor to release amino acids in the frozen potato samples, obtain a potato sample extraction solution, centrifuge the potato sample extraction solution, draw the supernatant from the potato sample extraction solution after centrifugation, activate a solid phase extraction column with methanol and water in sequence, pass the supernatant through the activated solid phase extraction column, remove the initial effluent to obtain an amino acid solution, elute the amino acid solution with an organic solvent eluent, collect the eluate, and dry the collected eluate on a nitrogen blower to obtain an amino acid sample in the potato;

[0016] The amino acid sample in the potato is processed to obtain a sample solution of the amino acid in the potato. The water transmission parameters of the amino acid sample solution in the potato are collected based on the distribution coefficient of different amino acids between the stationary phase and the mobile phase using a high performance liquid chromatograph; the pH value of the amino acid sample solution in the potato is collected using a pH meter; and the optical rotation angle of the amino acid sample solution in the potato is collected using a polarimeter.

[0017] A further improvement of the technical solution of the present invention is that: in step 1, the process of preprocessing the collected data includes:

[0018] The water transmission parameters, pH value and optical rotation angle of the collected amino acid sample solutions in potatoes were cleaned, and the average values of the water transmission parameters, pH value and optical rotation angle of the amino acid sample solutions in potatoes were calculated respectively. The corresponding calculated average values were used to fill the missing values in the water transmission parameters, pH value and optical rotation angle of the collected amino acid sample solutions.

[0019] A further improvement of the technical solution of the present invention is that in step 2, the process of obtaining the specific rotation of the amino acid sample solution in potatoes includes:

[0020] The optical rotation angle of the amino acid sample solution in potato is used as the first data set, and is divided into a training set and a test set in a ratio of 8:2. The amino acid specific optical rotation model of the first data set is used as input, and the specific optical rotation of the amino acid sample solution in potato is used as output. Using the training set data and a linear regression algorithm, the intercept term and the regression coefficient of the optical rotation angle of the amino acid sample solution in potato are adjusted to learn the linear relationship between the optical rotation angle of the amino acid sample solution in potato and the specific optical rotation of the amino acid sample solution in potato, thereby training the amino acid specific optical rotation model;

[0021] Input the test set data into the amino acid specific optical rotation model, evaluate the performance of the amino acid specific optical rotation model, adjust the parameters of the amino acid specific optical rotation model, optimize the amino acid specific optical rotation model, deploy the optimized amino acid specific optical rotation model into the system, and obtain the final amino acid specific optical rotation model;

[0022] The optical rotation angle of the amino acid sample solution in potato is input into the amino acid specific rotation model to obtain the specific rotation of the amino acid sample solution in potato.

[0023] A further improvement of the technical solution of the present invention is that in step 3, the process of determining the classification criteria of the amino acid sample solutions includes:

[0024] Obtaining historical data of amino acid samples in potatoes by searching PubMed database, wherein the historical data of amino acid samples in potatoes include historical water supply parameters, historical pH values, and historical specific rotations of amino acid sample solutions in potatoes;

[0025] Setting a water transmission parameter threshold of the amino acid sample solution in potatoes, classifying the amino acid sample solution in potatoes with a historical water transmission parameter higher than the water transmission parameter threshold as a hydrophobic amino acid sample solution, and classifying the amino acid sample solution in potatoes with a historical water transmission parameter lower than the water transmission parameter threshold as a non-hydrophobic amino acid sample solution, and obtaining a hydrophobicity classification result of the amino acid sample solution;

[0026] The amino acid sample solution in potatoes with a historical pH value less than 5 is classified as an acidic amino acid sample solution, the amino acid sample solution in potatoes with a historical pH value between 5 and 7 is classified as a neutral amino acid sample solution, and the amino acid sample solution in potatoes with a historical pH value greater than 7 is classified as an alkaline amino acid sample solution, thereby obtaining an acidity and alkalinity classification result of the amino acid sample solution;

[0027] The amino acid sample solution in potato with a historical specific optical rotation of 0 is divided into an achiral amino acid sample solution, and the amino acid sample solution in potato with a historical specific optical rotation not of 0 is divided into a chiral amino acid sample solution, and the optical rotation division result of the amino acid sample solution is obtained.

[0028] A further improvement of the technical solution of the present invention is that in step 4, the process of constructing a classification model for the amino acid sample solution and then obtaining the types of the amino acid sample solution in potatoes includes:

[0029] The water transfer parameters of the amino acid sample solutions in potatoes and the corresponding hydrophobicity classification criteria of the amino acid sample solutions, the pH values of the amino acid sample solutions in potatoes and the corresponding acidity and alkalinity classification criteria of the amino acid sample solutions, and the specific optical rotation of the amino acid sample solutions in potatoes and the corresponding optical rotation classification criteria of the amino acid sample solutions were used as the second data set and divided into a training set and a test set in a ratio of 8:2;

[0030] Using training set data and convolutional neural network algorithm, a convolutional neural network architecture is designed. The water delivery parameters of the amino acid sample solution in potatoes, the pH value of the amino acid sample solution in potatoes, and the specific optical rotation of the amino acid sample solution in potatoes are used as inputs, and the hydrophobicity partitioning results of the amino acid sample solution, the acidity and alkalinity partitioning results of the amino acid sample solution, and the optical rotation partitioning results of the amino acid sample solution are used as outputs. The output value of the convolutional neural network is calculated by forward propagation, and the gradient is calculated by backpropagation. The convolutional neural network model parameters are updated according to the gradient. The nonlinear relationship between the water delivery parameters of the amino acid sample solution in potatoes and the hydrophobicity partitioning results of the corresponding amino acid sample solution, the nonlinear relationship between the pH value of the amino acid sample solution in potatoes and the acidity and alkalinity partitioning results of the corresponding amino acid sample solution, and the nonlinear relationship between the specific optical rotation of the amino acid sample solution in potatoes and the optical rotation partitioning results of the corresponding amino acid sample solution are learned, and the amino acid sample solution type partitioning model is trained.

[0031] Input the test set data into the amino acid sample solution classification model, compare the output of the amino acid sample solution classification model with the types of amino acid sample solutions in actual potatoes, evaluate the performance of the amino acid sample solution classification model, adjust the parameters of the amino acid sample solution classification model, optimize the amino acid sample solution classification model, deploy the optimized amino acid sample solution classification model into the system, and obtain the final amino acid sample solution classification model;

[0032] The water delivery parameters, pH value and specific rotation of the amino acid sample solution in potatoes are input into the amino acid sample solution classification model, and the amino acid sample solution classification model is used to output the corresponding type of the amino acid sample solution in potatoes.

[0033] A further improvement of the technical solution of the present invention is that in step 5, the process of obtaining the liquid chromatography column distribution result includes:

[0034] For the hydrophobicity partitioning result of the amino acid sample solution, when the hydrophobicity partitioning result of the amino acid sample solution is a hydrophobic amino acid sample solution, a reverse phase liquid chromatography column is assigned; when the hydrophobicity partitioning result of the amino acid sample solution is a non-hydrophobic amino acid sample solution, a normal phase liquid chromatography column is assigned to obtain a hydrophobic liquid chromatography column assignment result;

[0035] For the acidity and alkalinity partition result of the amino acid sample solution, when the acidity and alkalinity partition result of the amino acid sample solution is an acidic amino acid sample solution, a strong cation exchange liquid chromatography column is assigned; when the acidity and alkalinity partition result of the amino acid sample solution is a neutral amino acid sample solution, a size exclusion liquid chromatography column is assigned; when the acidity and alkalinity partition result of the amino acid sample solution is a basic amino acid sample solution, a strong anion exchange liquid chromatography column is assigned to obtain the acidity and alkalinity liquid chromatography column assignment result;

[0036] For the optical rotation partition result of the amino acid sample solution, when the optical rotation partition result of the amino acid sample solution is a chiral amino acid sample solution, a chiral liquid chromatography column is assigned; when the optical rotation partition result of the amino acid sample solution is an achiral amino acid sample solution, a reversed-phase liquid chromatography column is assigned to obtain the optical rotation liquid chromatography column assignment result.

[0037] A further improvement of the technical solution of the present invention is that in step 5, the process of constructing the liquid chromatography column distribution model includes:

[0038] The hydrophobicity partitioning results of the amino acid sample solutions and their corresponding hydrophobicity liquid chromatography column assignment results, the acidity and alkalinity partitioning results of the amino acid sample solutions and their corresponding acidity and alkalinity liquid chromatography column assignment results, and the optical rotation partitioning results of the amino acid sample solutions and their corresponding optical rotation liquid chromatography column assignment results are used as the third data set and divided into a training set and a test set in a ratio of 7:3;

[0039] Combining the training set data with the neural network algorithm, the hydrophobicity partitioning results of the amino acid sample solution, the acidity and alkalinity partitioning results of the amino acid sample solution, and the optical rotation partitioning results of the amino acid sample solution are used as input, and the hydrophobicity liquid chromatography column distribution results, the acidity and alkalinity liquid chromatography column distribution results, and the optical rotation liquid chromatography column distribution results are used as output. By combining forward propagation and backpropagation, the nonlinear relationship between the hydrophobicity partitioning results of the amino acid sample solution and its corresponding hydrophobicity liquid chromatography column distribution results, the nonlinear relationship between the acidity and alkalinity liquid chromatography column distribution results of the amino acid sample solution, and the nonlinear relationship between the optical rotation partitioning results of the amino acid sample solution and its corresponding optical rotation liquid chromatography column distribution results are learned, thereby obtaining a trained liquid chromatography column distribution model;

[0040] The test set data is input into the trained liquid chromatography column allocation model. The MSE function is used to evaluate the error between the output results of the liquid chromatography column allocation model and the actual liquid chromatography column allocation results. According to the evaluation results, the parameters of the liquid chromatography column allocation model are adjusted to optimize the performance of the liquid chromatography column allocation model. The optimized liquid chromatography column allocation model is deployed into the system to obtain the liquid chromatography column allocation model.

[0041] A further improvement of the technical solution of the present invention is that in step five and step six, the process of obtaining the liquid chromatography column distribution result includes:

[0042] Inputting the hydrophobicity partitioning result, the acidity and alkalinity partitioning result, and the optical rotation partitioning result of the amino acid sample solution into the liquid chromatography column partitioning model, the liquid chromatography column partitioning model outputs the hydrophobicity liquid chromatography column partitioning result, the acidity and alkalinity liquid chromatography column partitioning result, and the optical rotation liquid chromatography column partitioning result, respectively;

[0043] When the liquid chromatography column allocation model outputs a single liquid chromatography column allocation result, the liquid chromatography column corresponding to the single liquid chromatography column allocation result is allocated to the amino acid sample solution in the potato;

[0044] When the liquid chromatography column allocation model outputs more than one liquid chromatography column allocation result, it indicates that the amino acid sample solution contains more than one chemical property at the same time. According to the liquid chromatography column allocation results output by the liquid chromatography column allocation model, the corresponding liquid chromatography columns are allocated in sequence to separate the amino acid sample solution in potatoes.

[0045] A further improvement of the technical solution of the present invention is that in step 6, the process of quantitative analysis of potato amino acids by liquid chromatography tandem mass spectrometry includes:

[0046] S1. Using an amino acid sample extracted from a potato, preparing amino acid sample solutions of different concentrations, the amino acid sample solutions including a target amino acid and an internal standard, collecting the concentration of the amino acid sample solutions using a fully automatic amino acid analyzer, detecting the chemical properties of the amino acid sample solutions, assigning a liquid chromatography column to the amino acid sample solutions according to the chemical properties of the amino acid sample solutions, setting the temperature of the liquid chromatography column to 30 degrees Celsius, setting a gradient according to the separation requirements of the amino acids, and separating amino acids of different chemical properties in the amino acid sample solutions;

[0047] S2. Using an electrospray ion source in positive ion mode, the spray voltage, capillary temperature, sheath gas flow rate, and auxiliary gas flow rate are set respectively. For each amino acid with different chemical properties, its parent ion and characteristic daughter ion are determined, and the corresponding monitoring ion pairs and collision energy are set. The amino acid solutions with different chemical properties flowing out of the liquid chromatography column are converted into gaseous ions;

[0048] S3. Using a liquid chromatography tandem mass spectrometer to receive gaseous ions generated by an electrospray ionization source, separating gaseous ions of different mass-to-charge ratios based on the motion characteristics of the gaseous ions in electric and magnetic fields, detecting the separated gaseous ions, and recording the peak areas of the target amino acid and the internal standard at each amino acid sample solution concentration;

[0049] S4. Draw a standard curve with the concentration of the amino acid sample solution as the horizontal axis and the peak area of the target amino acid and the internal standard as the vertical axis, and obtain the equation and correlation coefficient of the standard curve through linear regression analysis;

[0050] S5. Calculate the concentration of the amino acid in the amino acid sample solution based on the peak areas of the target amino acid and the internal standard and the equation of the standard curve.

[0051] The beneficial effects of the present invention are: a quantitative analysis method for potato amino acids by liquid chromatography-tandem mass spectrometry. Compared with the traditional quantitative analysis method for potato amino acids by liquid chromatography-tandem mass spectrometry, the linear regression algorithm, convolutional neural network algorithm and neural network algorithm in the method of the present invention are closely combined with modern information technology. Through a high-performance liquid chromatograph, a pH meter and a polarimeter, the chemical property data of the amino acid sample is accurately captured. The linear regression algorithm is used to obtain the specific rotation of the amino acid sample solution in the potato. The convolutional neural network algorithm is used to construct an amino acid sample solution type classification model to obtain the type of the amino acid sample solution in the potato, and then obtain the liquid chromatography column distribution result. The neural network algorithm is used to A liquid chromatography column distribution model is constructed, and a liquid chromatography column is distributed to an amino acid sample solution in a potato through the liquid chromatography column distribution model, thereby achieving the monitoring of the chemical properties of the amino acid sample. This solves the problem that multiple amino acids with similar chemical properties in potatoes have similar retention behaviors on the liquid chromatography column, resulting in overlapping chromatographic peaks and affecting the quantitative analysis results. This ensures that the method of the present invention can refine the dynamic monitoring standards for a quantitative analysis method for potato amino acids by liquid chromatography-tandem mass spectrometry within a more precise range, making the monitored data a more accurate indicator under the same conditions. The development and application of this method significantly enhance the intelligence level of the quantitative analysis method process for potato amino acids by liquid chromatography-tandem mass spectrometry. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a flow chart of a quantitative analysis method of potato amino acids by liquid chromatography-tandem mass spectrometry according to an embodiment of the present application. DETAILED DESCRIPTION

[0053] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0054] like Figure 1 As shown, the present invention provides a technical solution: a quantitative analysis method of potato amino acids by liquid chromatography tandem mass spectrometry, comprising the following steps:

[0055] Step 1: Obtain amino acid samples and their chemical property data from potatoes, and pre-process the acquired data to provide data support for subsequent steps;

[0056] Step 2: Based on the obtained optical rotation angle of the amino acid sample solution in potatoes, combined with a linear regression algorithm, the specific optical rotation of the amino acid sample solution in potatoes is obtained;

[0057] Step 3: Determine the classification criteria for amino acid sample solutions based on historical data of amino acid samples in potatoes, providing data support for constructing a classification model for amino acid sample solutions;

[0058] Step 4: Based on the classification criteria of the amino acid sample solutions, combined with the convolutional neural network algorithm, a classification model of the amino acid sample solutions is constructed to obtain the types of the amino acid sample solutions in potatoes;

[0059] Step 5: Obtaining liquid chromatography column partitioning results based on the types of amino acid sample solutions in potatoes, and constructing a liquid chromatography column partitioning model using the liquid chromatography column partitioning results. This solves the problem that potatoes contain multiple amino acids with similar chemical properties, resulting in overlapping chromatographic peaks and affecting the quantitative analysis results of potato amino acids by liquid chromatography-tandem mass spectrometry;

[0060] Step 6: Combine the output results of the liquid chromatography column distribution model to perform quantitative analysis of potato amino acids by liquid chromatography tandem mass spectrometry.

[0061] Preferably, in step 1, the process of obtaining the amino acid sample and the chemical property data of the amino acid sample includes:

[0062] Deploy different collection equipment to obtain amino acid samples and chemical property data from potatoes, wherein the collection equipment includes filter paper, centrifuge tubes, ultrasonic extractors, solid phase extraction columns, nitrogen blowdown devices, sample tubes, high performance liquid chromatographs, pH meters, and polarimeters;

[0063] The chemical property data of the amino acid sample include water delivery parameters, pH value and optical rotation angle of the amino acid sample solution in potatoes;

[0064] Select potatoes, rinse the potatoes with deionized water, dry the potatoes with filter paper, peel the potatoes and cut them into small pieces, freeze them in liquid nitrogen to obtain frozen potatoes, weigh the frozen potato samples, place the frozen potato samples in a centrifuge tube filled with an extraction solvent, place the centrifuge tube in an ultrasonic extractor to release amino acids in the frozen potato samples, obtain a potato sample extraction solution, centrifuge the potato sample extraction solution, draw the supernatant from the potato sample extraction solution after centrifugation, activate a solid phase extraction column with methanol and water in sequence, pass the supernatant through the activated solid phase extraction column, remove the initial effluent to obtain an amino acid solution, elute the amino acid solution with an organic solvent eluent, collect the eluate, and dry the collected eluate on a nitrogen blower to obtain an amino acid sample in the potato;

[0065] The collected amino acid sample in the potato is placed in a sample tube, clean water is added to the sample tube, and the mixture is stirred to obtain a sample solution of the amino acid in the potato. The water transmission parameters of the amino acid sample solution in the potato are collected using a high performance liquid chromatograph based on the distribution coefficient of different amino acids between the stationary phase and the mobile phase; the pH value of the amino acid sample solution in the potato is collected using a pH meter; and the optical rotation angle of the amino acid sample solution in the potato is collected using a polarimeter.

[0066] Preferably, in step 1, the process of preprocessing the collected data includes:

[0067] The water transmission parameters, pH value and optical rotation angle of the collected amino acid sample solutions in potatoes were cleaned, and the average values of the water transmission parameters, pH value and optical rotation angle of the amino acid sample solutions in potatoes were calculated respectively. The corresponding calculated average values were used to fill the missing values in the water transmission parameters, pH value and optical rotation angle of the collected amino acid sample solutions.

[0068] Preferably, in step 2, the process of obtaining the specific rotation of the amino acid sample solution in potatoes includes:

[0069] The optical rotation angle of the amino acid sample solution in potato is used as the first data set, and is divided into a training set and a test set in a ratio of 8:2. The amino acid specific optical rotation model of the first data set is used as input, and the specific optical rotation of the amino acid sample solution in potato is used as output. Using the training set data and a linear regression algorithm, the intercept term and the regression coefficient of the optical rotation angle of the amino acid sample solution in potato are adjusted to learn the linear relationship between the optical rotation angle of the amino acid sample solution in potato and the specific optical rotation of the amino acid sample solution in potato, thereby training the amino acid specific optical rotation model;

[0070] Input the test set data into the amino acid specific optical rotation model, evaluate the performance of the amino acid specific optical rotation model, adjust the parameters of the amino acid specific optical rotation model, optimize the amino acid specific optical rotation model, deploy the optimized amino acid specific optical rotation model into the system, and obtain the final amino acid specific optical rotation model;

[0071] The optical rotation angle of the amino acid sample solution in potato is input into the amino acid specific rotation model to obtain the specific rotation of the amino acid sample solution in potato.

[0072] Preferably, in step 3, the process of determining the classification criteria of the amino acid sample solutions includes:

[0073] Obtaining historical data of amino acid samples in potatoes by searching PubMed database, wherein the historical data of amino acid samples in potatoes include historical water supply parameters, historical pH values, and historical specific rotations of amino acid sample solutions in potatoes;

[0074] Setting a water transmission parameter threshold of the amino acid sample solution in potatoes, classifying the amino acid sample solution in potatoes with a historical water transmission parameter higher than the water transmission parameter threshold as a hydrophobic amino acid sample solution, and classifying the amino acid sample solution in potatoes with a historical water transmission parameter lower than the water transmission parameter threshold as a non-hydrophobic amino acid sample solution, and obtaining a hydrophobicity classification result of the amino acid sample solution;

[0075] The amino acid sample solution in potatoes with a historical pH value less than 5 is classified as an acidic amino acid sample solution, the amino acid sample solution in potatoes with a historical pH value between 5 and 7 is classified as a neutral amino acid sample solution, and the amino acid sample solution in potatoes with a historical pH value greater than 7 is classified as an alkaline amino acid sample solution, thereby obtaining an acidity and alkalinity classification result of the amino acid sample solution;

[0076] The amino acid sample solution in potato with a historical specific optical rotation of 0 is divided into an achiral amino acid sample solution, and the amino acid sample solution in potato with a historical specific optical rotation not of 0 is divided into a chiral amino acid sample solution, and the optical rotation division result of the amino acid sample solution is obtained.

[0077] Preferably, in step 4, the process of constructing a classification model of amino acid sample solutions to obtain the types of amino acid sample solutions in potatoes includes:

[0078] The water transfer parameters of the amino acid sample solutions in potatoes and the corresponding hydrophobicity classification criteria of the amino acid sample solutions, the pH values of the amino acid sample solutions in potatoes and the corresponding acidity and alkalinity classification criteria of the amino acid sample solutions, and the specific optical rotation of the amino acid sample solutions in potatoes and the corresponding optical rotation classification criteria of the amino acid sample solutions were used as the second data set and divided into a training set and a test set in a ratio of 8:2;

[0079] Using training set data and convolutional neural network algorithm, a convolutional neural network architecture is designed. The water delivery parameters of the amino acid sample solution in potatoes, the pH value of the amino acid sample solution in potatoes, and the specific optical rotation of the amino acid sample solution in potatoes are used as inputs, and the hydrophobicity partitioning results of the amino acid sample solution, the acidity and alkalinity partitioning results of the amino acid sample solution, and the optical rotation partitioning results of the amino acid sample solution are used as outputs. The output value of the convolutional neural network is calculated by forward propagation, and the gradient is calculated by backpropagation. The convolutional neural network model parameters are updated according to the gradient. The nonlinear relationship between the water delivery parameters of the amino acid sample solution in potatoes and the hydrophobicity partitioning results of the corresponding amino acid sample solution, the nonlinear relationship between the pH value of the amino acid sample solution in potatoes and the acidity and alkalinity partitioning results of the corresponding amino acid sample solution, and the nonlinear relationship between the specific optical rotation of the amino acid sample solution in potatoes and the optical rotation partitioning results of the corresponding amino acid sample solution are learned, and the amino acid sample solution type partitioning model is trained.

[0080] Input the test set data into the amino acid sample solution classification model, compare the output of the amino acid sample solution classification model with the types of amino acid sample solutions in actual potatoes, evaluate the performance of the amino acid sample solution classification model, adjust the parameters of the amino acid sample solution classification model, optimize the amino acid sample solution classification model, deploy the optimized amino acid sample solution classification model into the system, and obtain the final amino acid sample solution classification model;

[0081] The water delivery parameters, pH value and specific rotation of the amino acid sample solution in potatoes are input into the amino acid sample solution classification model, and the amino acid sample solution classification model is used to output the corresponding type of the amino acid sample solution in potatoes.

[0082] Preferably, in step 5, the process of obtaining the liquid chromatography column distribution result includes:

[0083] For the hydrophobicity partitioning result of the amino acid sample solution, when the hydrophobicity partitioning result of the amino acid sample solution is a hydrophobic amino acid sample solution, a reverse phase liquid chromatography column is assigned; when the hydrophobicity partitioning result of the amino acid sample solution is a non-hydrophobic amino acid sample solution, a normal phase liquid chromatography column is assigned to obtain a hydrophobic liquid chromatography column assignment result;

[0084] For the acidity and alkalinity partition result of the amino acid sample solution, when the acidity and alkalinity partition result of the amino acid sample solution is an acidic amino acid sample solution, a strong cation exchange liquid chromatography column is assigned; when the acidity and alkalinity partition result of the amino acid sample solution is a neutral amino acid sample solution, a size exclusion liquid chromatography column is assigned; when the acidity and alkalinity partition result of the amino acid sample solution is a basic amino acid sample solution, a strong anion exchange liquid chromatography column is assigned to obtain the acidity and alkalinity liquid chromatography column assignment result;

[0085] For the optical rotation partition result of the amino acid sample solution, when the optical rotation partition result of the amino acid sample solution is a chiral amino acid sample solution, a chiral liquid chromatography column is assigned; when the optical rotation partition result of the amino acid sample solution is an achiral amino acid sample solution, a reversed-phase liquid chromatography column is assigned to obtain the optical rotation liquid chromatography column assignment result.

[0086] Preferably, in step 5, the process of constructing the liquid chromatography column distribution model includes:

[0087] The hydrophobicity partitioning results of the amino acid sample solutions and their corresponding hydrophobicity liquid chromatography column assignment results, the acidity and alkalinity partitioning results of the amino acid sample solutions and their corresponding acidity and alkalinity liquid chromatography column assignment results, and the optical rotation partitioning results of the amino acid sample solutions and their corresponding optical rotation liquid chromatography column assignment results are used as the third data set and divided into a training set and a test set in a ratio of 7:3;

[0088] Combining the training set data with the neural network algorithm, the hydrophobicity partitioning results of the amino acid sample solution, the acidity and alkalinity partitioning results of the amino acid sample solution, and the optical rotation partitioning results of the amino acid sample solution are used as input, and the hydrophobicity liquid chromatography column distribution results, the acidity and alkalinity liquid chromatography column distribution results, and the optical rotation liquid chromatography column distribution results are used as output. By combining forward propagation and backpropagation, the nonlinear relationship between the hydrophobicity partitioning results of the amino acid sample solution and its corresponding hydrophobicity liquid chromatography column distribution results, the nonlinear relationship between the acidity and alkalinity liquid chromatography column distribution results of the amino acid sample solution, and the nonlinear relationship between the optical rotation partitioning results of the amino acid sample solution and its corresponding optical rotation liquid chromatography column distribution results are learned, thereby obtaining a trained liquid chromatography column distribution model;

[0089] The test set data is input into the trained liquid chromatography column allocation model. The MSE function is used to evaluate the error between the output results of the liquid chromatography column allocation model and the actual liquid chromatography column allocation results. According to the evaluation results, the parameters of the liquid chromatography column allocation model are adjusted to optimize the performance of the liquid chromatography column allocation model. The optimized liquid chromatography column allocation model is deployed into the system to obtain the liquid chromatography column allocation model.

[0090] Preferably, in step 6, the process of obtaining the liquid chromatography column distribution result includes:

[0091] Inputting the hydrophobicity partitioning result, the acidity and alkalinity partitioning result, and the optical rotation partitioning result of the amino acid sample solution into the liquid chromatography column partitioning model, the liquid chromatography column partitioning model outputs the hydrophobicity liquid chromatography column partitioning result, the acidity and alkalinity liquid chromatography column partitioning result, and the optical rotation liquid chromatography column partitioning result, respectively;

[0092] When the liquid chromatography column allocation model outputs a single liquid chromatography column allocation result, the liquid chromatography column corresponding to the single liquid chromatography column allocation result is allocated to the amino acid sample solution in the potato;

[0093] When the liquid chromatography column allocation model outputs more than one liquid chromatography column allocation result, it indicates that the amino acid sample solution contains more than one chemical property at the same time. According to the liquid chromatography column allocation results output by the liquid chromatography column allocation model, the corresponding liquid chromatography columns are allocated in sequence to separate the amino acid sample solution in potatoes.

[0094] Preferably, in step six, the process of performing quantitative analysis of potato amino acids by liquid chromatography tandem mass spectrometry includes:

[0095] S1. Using an amino acid sample extracted from a potato, preparing amino acid sample solutions of different concentrations, the amino acid sample solutions including a target amino acid and an internal standard, collecting the concentration of the amino acid sample solutions using a fully automatic amino acid analyzer, detecting the chemical properties of the amino acid sample solutions, assigning a liquid chromatography column to the amino acid sample solutions according to the chemical properties of the amino acid sample solutions, setting the temperature of the liquid chromatography column to 30 degrees Celsius, setting a gradient according to the separation requirements of the amino acids, and separating amino acids of different chemical properties in the amino acid sample solutions;

[0096] S2. Using an electrospray ion source in positive ion mode, the spray voltage, capillary temperature, sheath gas flow rate, and auxiliary gas flow rate are set respectively. For each amino acid with different chemical properties, its parent ion and characteristic daughter ion are determined, and the corresponding monitoring ion pairs and collision energy are set. The amino acid solutions with different chemical properties flowing out of the liquid chromatography column are converted into gaseous ions;

[0097] S3. Using a liquid chromatography tandem mass spectrometer to receive gaseous ions generated by an electrospray ionization source, separating gaseous ions of different mass-to-charge ratios based on the motion characteristics of the gaseous ions in electric and magnetic fields, detecting the separated gaseous ions, and recording the peak areas of the target amino acid and the internal standard at each amino acid sample solution concentration;

[0098] S4. Draw a standard curve with the concentration of the amino acid sample solution as the horizontal axis and the peak area of the target amino acid and the internal standard as the vertical axis, and obtain the equation and correlation coefficient of the standard curve through linear regression analysis;

[0099] S5. Calculate the concentration of the amino acid in the amino acid sample solution based on the peak areas of the target amino acid and the internal standard and the equation of the standard curve.

[0100] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "include," "comprise," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations. The phrase "includes an element defined by..." does not exclude the presence of other identical elements in the process, method, article, or device that includes the element.

[0101] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A quantitative analysis method for potato amino acids by liquid chromatography-tandem mass spectrometry, characterized in that: The following steps are involved: Step 1: Obtain amino acid samples and chemical property data from potatoes, and pre-process the obtained data; in step 1, the process of pre-processing the obtained data includes: Deploy various collection equipment to obtain amino acid samples and their chemical property data from potatoes, including filter paper, centrifuge tubes, ultrasonic extractors, solid-phase extraction columns, nitrogen blowdown devices, sample tubes, high-performance liquid chromatographs, pH meters, and polarimeters. Data cleaning is performed on the water delivery parameters, pH values, and optical rotation angles of the collected amino acid sample solutions from potatoes. Average values of the water delivery parameters, pH values, and optical rotation angles of the amino acid sample solutions from potatoes are calculated, and the corresponding average values are used to fill in missing values in the water delivery parameters, pH values, and optical rotation angles of the collected amino acid sample solutions. Step 2: Based on the pretreated amino acid sample in the potato and its chemical property data, combined with a linear regression algorithm, the specific rotation of the amino acid sample solution in the potato is obtained; Step 3: Determine the classification criteria for the amino acid sample solutions based on the historical data of the amino acid samples in potatoes. In step 3, determining the classification criteria for the amino acid sample solutions includes: obtaining the historical data of the amino acid samples in potatoes by searching the PubMed database, wherein the historical data of the amino acid samples in potatoes include historical water supply parameters, historical pH values, and historical specific rotations of the amino acid sample solutions in potatoes; Setting a water transmission parameter threshold of the amino acid sample solution in potatoes, classifying the amino acid sample solution in potatoes with a historical water transmission parameter higher than the water transmission parameter threshold as a hydrophobic amino acid sample solution, and classifying the amino acid sample solution in potatoes with a historical water transmission parameter lower than the water transmission parameter threshold as a non-hydrophobic amino acid sample solution, and obtaining a hydrophobicity classification result of the amino acid sample solution; The amino acid sample solution in potatoes with a historical pH value of less than 5 is classified as an acidic amino acid sample solution, the amino acid sample solution in potatoes with a historical pH value between 5 and 7 is classified as a neutral amino acid sample solution, and the amino acid sample solution in potatoes with a historical pH value greater than 7 is classified as an alkaline amino acid sample solution, thereby obtaining the acidity and alkalinity classification results of the amino acid sample solutions; dividing the amino acid sample solution in potatoes with a historical specific optical rotation of 0 into an achiral amino acid sample solution, and dividing the amino acid sample solution in potatoes with a historical specific optical rotation not equal to 0 into a chiral amino acid sample solution, and obtaining an optical rotation division result of the amino acid sample solution; Step 4: Based on the classification criteria of the amino acid sample solution types, combined with the convolutional neural network algorithm, a model for classifying the amino acid sample solution types is constructed to obtain the types of the amino acid sample solution in the potato. In step 4, the process of constructing the model for classifying the amino acid sample solution types to obtain the types of the amino acid sample solution in the potato includes: The water transfer parameters of the amino acid sample solutions in potatoes and their corresponding hydrophobicity classification criteria of the amino acid sample solutions, the pH values of the amino acid sample solutions in potatoes and their corresponding acidity and alkalinity classification criteria of the amino acid sample solutions, and the specific optical rotations of the amino acid sample solutions in potatoes and their corresponding optical rotation classification criteria of the amino acid sample solutions were used as the second data set and divided into a training set and a test set in a ratio of 8:2; Using training set data and a convolutional neural network algorithm, a convolutional neural network architecture is designed. The water delivery parameters, pH value, and specific optical rotation of the amino acid sample solution in potatoes are used as inputs, and the hydrophobicity partitioning results, acidity-alkalinity partitioning results, and optical rotation partitioning results of the amino acid sample solution are used as outputs. The output values of the convolutional neural network are calculated through forward propagation, and the gradients are calculated through backpropagation. The convolutional neural network model parameters are updated according to the gradients. The nonlinear relationships between the water delivery parameters of the amino acid sample solution in potatoes and the hydrophobicity partitioning results of the corresponding amino acid sample solutions, the nonlinear relationships between the pH value of the amino acid sample solution in potatoes and the acidity-alkalinity partitioning results of the corresponding amino acid sample solutions, and the nonlinear relationships between the specific optical rotation of the amino acid sample solution in potatoes and the optical rotation partitioning results of the corresponding amino acid sample solutions are learned, and a model for classifying the amino acid sample solutions is trained. Input the test set data into the amino acid sample solution classification model, compare the output of the amino acid sample solution classification model with the types of amino acid sample solutions in actual potatoes, evaluate the performance of the amino acid sample solution classification model, adjust the parameters of the amino acid sample solution classification model, optimize the amino acid sample solution classification model, deploy the optimized amino acid sample solution classification model into the system, and obtain the final amino acid sample solution classification model; Step 5: Obtaining liquid chromatography column distribution results according to the types of amino acid sample solutions in potatoes, and constructing a liquid chromatography column distribution model using the liquid chromatography column distribution results; in step 5, the process of constructing the liquid chromatography column distribution model includes: The hydrophobicity partitioning results of the amino acid sample solutions and their corresponding hydrophobicity liquid chromatography column assignment results, the acidity and alkalinity partitioning results of the amino acid sample solutions and their corresponding acidity and alkalinity liquid chromatography column assignment results, and the optical rotation partitioning results of the amino acid sample solutions and their corresponding optical rotation liquid chromatography column assignment results are used as the third data set and divided into a training set and a test set in a ratio of 7:3; Combining training set data with a neural network algorithm, the hydrophobicity partitioning results of the amino acid sample solution, the acidity and alkalinity partitioning results of the amino acid sample solution, and the optical rotation partitioning results of the amino acid sample solution are used as input, and the hydrophobicity liquid chromatography column assignment results, the acidity and alkalinity liquid chromatography column assignment results, and the optical rotation liquid chromatography column assignment results are used as output. By combining forward propagation and backpropagation, the nonlinear relationship between the hydrophobicity partitioning results of the amino acid sample solution and its corresponding hydrophobicity liquid chromatography column assignment results, the nonlinear relationship between the acidity and alkalinity liquid chromatography column assignment results of the amino acid sample solution, and the nonlinear relationship between the optical rotation partitioning results of the amino acid sample solution and its corresponding optical rotation liquid chromatography column assignment results are learned, thereby training a liquid chromatography column assignment model; Step 6: Combine the output results of the liquid chromatography column distribution model to perform quantitative analysis of potato amino acids by liquid chromatography tandem mass spectrometry.

2. The quantitative analysis method of potato amino acids by liquid chromatography-tandem mass spectrometry according to claim 1, characterized in that: In step 1, the process of obtaining amino acid samples and chemical property data in potatoes includes: The chemical property data of the amino acid sample include water transfer parameters, pH value and optical rotation angle of the amino acid sample solution in potatoes.

3. The quantitative analysis method of potato amino acids by liquid chromatography-tandem mass spectrometry according to claim 2, characterized in that: In the step 2, the process of obtaining the specific rotation of the amino acid sample solution in potatoes includes: The optical rotation angle of the amino acid sample solution in potato is used as the first data set, and is divided into a training set and a test set in a ratio of 8:

2. The amino acid specific optical rotation model of the first data set is used as input, and the specific optical rotation of the amino acid sample solution in potato is used as output. Using the training set data and a linear regression algorithm, the intercept term and the regression coefficient of the optical rotation angle of the amino acid sample solution in potato are adjusted to learn the linear relationship between the optical rotation angle of the amino acid sample solution in potato and the specific optical rotation of the amino acid sample solution in potato, thereby training the amino acid specific optical rotation model; Input the test set data into the amino acid specific optical rotation model, evaluate the performance of the amino acid specific optical rotation model, adjust the parameters of the amino acid specific optical rotation model, optimize the amino acid specific optical rotation model, deploy the optimized amino acid specific optical rotation model into the system, and obtain the final amino acid specific optical rotation model; The optical rotation angle of the amino acid sample solution in potato is input into the amino acid specific rotation model to obtain the specific rotation of the amino acid sample solution in potato.

4. The quantitative analysis method of potato amino acids by liquid chromatography-tandem mass spectrometry according to claim 3, characterized in that: In step 4, the process of constructing a classification model for the amino acid sample solution and then obtaining the types of the amino acid sample solution in potatoes includes: The water delivery parameters, pH value and specific rotation of the amino acid sample solution in potatoes are input into the amino acid sample solution classification model, and the amino acid sample solution classification model is used to output the corresponding type of the amino acid sample solution in potatoes.

5. The quantitative analysis method of potato amino acids by liquid chromatography tandem mass spectrometry according to claim 4, characterized in that: In step 5, the process of obtaining the liquid chromatography column distribution result includes: For the hydrophobicity partitioning result of the amino acid sample solution, when the hydrophobicity partitioning result of the amino acid sample solution is a hydrophobic amino acid sample solution, a reverse phase liquid chromatography column is assigned; when the hydrophobicity partitioning result of the amino acid sample solution is a non-hydrophobic amino acid sample solution, a normal phase liquid chromatography column is assigned to obtain a hydrophobic liquid chromatography column assignment result; For the acidity and alkalinity partition result of the amino acid sample solution, when the acidity and alkalinity partition result of the amino acid sample solution is an acidic amino acid sample solution, a strong cation exchange liquid chromatography column is assigned; when the acidity and alkalinity partition result of the amino acid sample solution is a neutral amino acid sample solution, a size exclusion liquid chromatography column is assigned; when the acidity and alkalinity partition result of the amino acid sample solution is a basic amino acid sample solution, a strong anion exchange liquid chromatography column is assigned to obtain the acidity and alkalinity liquid chromatography column assignment result; For the optical rotation partition result of the amino acid sample solution, when the optical rotation partition result of the amino acid sample solution is a chiral amino acid sample solution, a chiral liquid chromatography column is assigned; when the optical rotation partition result of the amino acid sample solution is an achiral amino acid sample solution, a reversed-phase liquid chromatography column is assigned to obtain the optical rotation liquid chromatography column assignment result.

6. The quantitative analysis method of potato amino acids by liquid chromatography tandem mass spectrometry according to claim 5, characterized in that: In the step 5, the process of constructing the liquid chromatography column distribution model includes: The test set data is input into the liquid chromatography column allocation model. The MSE function is used to evaluate the error between the output results of the liquid chromatography column allocation model and the actual liquid chromatography column allocation results. According to the evaluation results, the parameters of the liquid chromatography column allocation model are adjusted to optimize the performance of the liquid chromatography column allocation model. The optimized liquid chromatography column allocation model is deployed into the system to obtain the final liquid chromatography column allocation model.

7. The quantitative analysis method of potato amino acids by liquid chromatography tandem mass spectrometry according to claim 6, characterized in that: In step 6, the process of obtaining the liquid chromatography column distribution result includes: Inputting the hydrophobicity partitioning result, the acidity and alkalinity partitioning result, and the optical rotation partitioning result of the amino acid sample solution into the liquid chromatography column partitioning model, the liquid chromatography column partitioning model outputs the hydrophobicity liquid chromatography column partitioning result, the acidity and alkalinity liquid chromatography column partitioning result, and the optical rotation liquid chromatography column partitioning result, respectively; When the liquid chromatography column allocation model outputs a single liquid chromatography column allocation result, the liquid chromatography column corresponding to the single liquid chromatography column allocation result is allocated to the amino acid sample solution in the potato; When the liquid chromatography column allocation model outputs more than one liquid chromatography column allocation result, it indicates that the amino acid sample solution contains more than one chemical property at the same time. According to the liquid chromatography column allocation results output by the liquid chromatography column allocation model, the corresponding liquid chromatography columns are allocated in sequence to separate the amino acid sample solution in potatoes.

8. The quantitative analysis method of potato amino acids by liquid chromatography tandem mass spectrometry according to claim 7, characterized in that: In step 6, the process of quantitative analysis of potato amino acids by liquid chromatography tandem mass spectrometry includes: S1. Using an amino acid sample extracted from a potato, preparing amino acid sample solutions of different concentrations, the amino acid sample solutions including a target amino acid and an internal standard, collecting the concentration of the amino acid sample solutions using a fully automatic amino acid analyzer, detecting the chemical properties of the amino acid sample solutions, assigning a liquid chromatography column to the amino acid sample solutions according to the chemical properties of the amino acid sample solutions, setting the temperature of the liquid chromatography column to 30 degrees Celsius, setting a gradient according to the separation requirements of the amino acids, and separating amino acids of different chemical properties in the amino acid sample solutions; S2. Using an electrospray ion source in positive ion mode, the spray voltage, capillary temperature, sheath gas flow rate, and auxiliary gas flow rate are set respectively. For each amino acid with different chemical properties, its parent ion and characteristic daughter ion are determined, and the corresponding monitoring ion pairs and collision energy are set. The amino acid solutions with different chemical properties flowing out of the liquid chromatography column are converted into gaseous ions; S3. Using a liquid chromatography tandem mass spectrometer to receive gaseous ions generated by an electrospray ionization source, separating gaseous ions of different mass-to-charge ratios based on the motion characteristics of the gaseous ions in electric and magnetic fields, detecting the separated gaseous ions, and recording the peak areas of the target amino acid and the internal standard at each amino acid sample solution concentration; S4. Draw a standard curve with the concentration of the amino acid sample solution as the horizontal axis and the peak area of the target amino acid and the internal standard as the vertical axis, and obtain the equation and correlation coefficient of the standard curve through linear regression analysis; S5. Calculate the concentration of the amino acid in the amino acid sample solution based on the peak areas of the target amino acid and the internal standard and the equation of the standard curve.

Images