Preparation method of oxidized low-density lipoprotein quality control product prepared based on novel antigen

Abstract

The invention discloses a preparation method of an oxidized low-density lipoprotein quality control product prepared on the basis of a novel antigen, and the method comprises the following steps: constructing an oxidized low-density lipoprotein model, carrying out molecular dynamics simulation by setting a plurality of groups of simulation conditions to calculate a stability index, and determining an optimal antigen preparation condition; acquiring antigen epitope data, and constructing a feature vector, so as to predict and screen an optimal antigen epitope by utilizing a machine learning algorithm; and based on a target cross-linking strategy, combining the oxidized low-density lipoprotein prepared under the optimal antigen preparation condition with the optimal epitope to obtain a target quality control product. According to the scheme, the preparation efficiency and quality can be improved.

Description

Technical Field

[0001] The present application relates to the field of bioinformatics, and in particular to a method for preparing an oxidized low-density lipoprotein control product based on a novel antigen. Background Art

[0002] Oxidized low-density lipoprotein (oxLDL) plays a key role in the development of atherosclerosis and cardiovascular disease. As an important biomarker, accurate detection of oxLDL is of great significance for early diagnosis, risk assessment and treatment monitoring of cardiovascular diseases. However, the detection of oxLDL faces challenges in standardization and quality control, which is mainly due to the instability and heterogeneity of oxLDL quality control preparation. The preparation of oxLDL quality control is a complex process involving multiple steps such as separation, oxidative modification and stabilization of low-density lipoprotein (LDL). Traditional methods for preparing oxLDL quality control mainly rely on in vitro chemical oxidation or enzymatic oxidation. Although these methods can produce oxLDL, it is often difficult to accurately control the degree of oxidation and the uniformity of the product. In addition, due to the instability of oxLDL, the prepared quality control is prone to further oxidation or degradation during storage and use, resulting in inconsistency and reduced comparability of the test results. These problems seriously affect the accuracy and reliability of oxLDL detection and limit its application in clinical diagnosis and research.

[0003] In order to solve the problems existing in the preparation of oxLDL quality control products, researchers have tried a variety of improvement methods. These include optimizing oxidation conditions, using antioxidants to stabilize products, and developing new oxidation modification strategies. For example, some studies have obtained LDL with different degrees of oxidation by controlling the oxidation time and oxidant concentration to simulate oxLDL at different stages in the body. Other researchers have tried to use natural antioxidants or synthetic stabilizers to extend the shelf-life of oxLDL quality control products. There are also some innovative methods, such as using specific enzyme systems for directed oxidation, or preparing recombinant proteins that simulate oxLDL through genetic engineering technology. These methods have improved the stability and uniformity of oxLDL quality control products to a certain extent, but there are still some limitations. For example, the optimization of oxidation conditions often relies on an empirical trial and error process, which is difficult to accurately control; the use of antioxidants may affect subsequent detection; and although genetic engineering methods can produce highly uniform products, they may not be able to fully simulate the complex structure and modification of natural oxLDL. Compared with methods based on the preparation of new antigens, these existing technologies still have deficiencies in precise control and repeatability, and it is difficult to meet the needs of high-precision oxLDL detection.

[0004] Therefore, a technical solution is urgently needed to improve the preparation efficiency and quality. Summary of the invention

[0005] In order to solve the deficiencies of the prior art, the present application provides a method for preparing an oxidized low-density lipoprotein control product based on a novel antigen. The present application solves the technical problems that the prior art still has deficiencies in precise control and repeatability.

[0006] The embodiment of the present application provides a method for preparing an oxidized low-density lipoprotein control product based on a novel antigen, comprising: constructing an oxidized low-density lipoprotein model, performing molecular dynamics simulation by setting multiple groups of simulation conditions to calculate a stability index, and determining optimal antigen preparation conditions; obtaining antigen epitope data, and constructing a feature vector to predict and screen the optimal antigen epitope using a machine learning algorithm; based on a target cross-linking strategy, combining the oxidized low-density lipoprotein prepared under the optimal antigen preparation conditions with the optimal antigen epitope to obtain a target quality control product.

[0007] In a possible implementation, a model of oxidized low-density lipoprotein is constructed, and molecular dynamics simulation is performed by setting multiple groups of simulation conditions to calculate the stability index and determine the optimal antigen preparation conditions, including: constructing a molecular model of oxidized low-density lipoprotein, including obtaining an initial structure based on low-density lipoprotein structural modification, and parameterizing the initial structure; setting multiple groups of simulation conditions, including pH value, temperature and ion concentration, and constructing multiple simulation systems based on the simulation conditions; performing energy minimization, equilibrium and production simulation on the multiple simulation systems, respectively, to obtain multiple groups of simulation trajectories; analyzing the multiple groups of simulation trajectories, calculating the root mean square deviation and the number of hydrogen bonds; calculating the stability index based on the root mean square deviation, the number of hydrogen bonds and the energy fluctuation; comparing the stability index under different simulation conditions, and selecting the condition with the highest stability index as the optimal antigen preparation condition.

[0008] In a possible implementation, antigen epitope data is obtained and a feature vector is constructed to predict and screen the best antigen epitope using a machine learning algorithm, including: obtaining known oxidized low-density lipoprotein antigen epitope data from an immune epitope database, including amino acid sequences and their structural characteristics, using bioinformatics tools to process the sequence data and calculate secondary structure information; constructing a feature vector representing a known epitope, including amino acid composition, secondary structure composition, hydrophobicity index, isoelectric point, molecular weight and surface accessibility; based on the constructed feature vector, using an ensemble learning method to train a machine learning model to predict the epitope, and sort and screen the prediction results; according to the screened candidate epitopes, using solid phase synthesis or recombinant protein expression to prepare the corresponding synthetic peptide or recombinant protein.

[0009] In a possible implementation, based on a target cross-linking strategy, oxidized low-density lipoprotein prepared under optimal antigen preparation conditions is combined with an optimal antigen epitope to obtain a target quality control product, including: preparing oxidized low-density lipoprotein under preset pH, temperature and ionic strength conditions according to the results of the optimal antigen preparation conditions; preparing antigen epitope peptides by solid phase synthesis based on the optimal antigen epitope predicted by machine learning, and verifying the peptides by high performance liquid chromatography purification and mass spectrometry analysis; combining the prepared oxidized low-density lipoprotein with the synthesized antigen epitope peptides through a target chemical cross-linking reaction to obtain a quality control product.

[0010] In a possible implementation, calculating the stability index includes: ,in, represents the stability index, represents the root mean square deviation, represents the average number of hydrogen bonds, represents the theoretical maximum number of hydrogen bonds, represents the energy fluctuation of the system, represents the Boltzmann constant, Indicates absolute temperature.

[0011] In a possible implementation, an ensemble learning method is used to train a machine learning model based on the constructed feature vector to predict epitopes, and the prediction results are sorted and screened, including: based on the constructed feature vector, an ensemble learning method is used to train a machine learning model, and grid search and cross-validation are used to optimize the hyperparameters of the model; a sliding window scan is performed on the complete sequence of oxidized low-density lipoprotein, the feature vector of each fragment is calculated, and the probability of it being an epitope is predicted using the trained machine learning model; an epitope scoring function is constructed, the predicted epitope probability is combined with the relative solvent accessible area and the shortest distance from the center of the fragment to the protein surface, and a genetic algorithm is used to optimize the weight coefficient of the function; and the predicted epitopes are sorted and screened according to the constructed epitope scoring function.

[0012] In a possible implementation, the prepared oxidized low-density lipoprotein is combined with the synthesized antigen epitope peptide segment through a targeted chemical cross-linking reaction, comprising: ,in, stands for N-hydroxysuccinimide activated oxidized low-density lipoprotein, Indicates the epitope peptide with cysteine ​​added to the C-terminus. Represents a cross-linked product, Denotes N-hydroxysuccinimide, a by-product.

[0013] In a possible implementation, after the quality control product is prepared, the method further comprises: verifying the specificity of the quality control product by enzyme-linked immunosorbent assay, and evaluating its stability by introducing a stability factor; wherein the stability factor comprises: ,in, represents the stability factor, represents the ELISA signal intensity after storage for t days, represents the initial ELISA signal intensity, Represents the attenuation constant in units of , Indicates storage time, Indicates the difference in denaturation temperature before and after storage. represents the initial denaturation temperature.

[0014] In the method for preparing an oxidized low-density lipoprotein control product based on a novel antigen as provided above, the embodiment of the present application can reduce the experimental trial and error cost by optimizing the antigen preparation conditions through molecular dynamics simulation, and further, by applying machine learning algorithms to predict and screen the optimal antigen epitopes, it can improve the specificity of the quality control product as well as the preparation efficiency and quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0016] Figure 1 A schematic diagram of a process for preparing an oxidized low-density lipoprotein control product based on a novel antigen provided in an embodiment of the present application; Figure 2 A schematic diagram of a method for optimizing antigen preparation conditions using molecular dynamics simulation provided in an embodiment of the present application; Figure 3 A schematic diagram of a process for predicting and screening optimal antigen epitopes using a machine learning algorithm provided in an embodiment of the present application; Figure 4 A schematic diagram of a process for preparing a quality control product provided in an embodiment of the present application; Figure 5 A schematic diagram of a molecular dynamics simulation result provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present application.

[0018] Those skilled in the art will appreciate that the terms "first", "second" and the like in the embodiments of the present application are only used to distinguish different steps, devices or modules, etc., and do not represent any specific technical meaning, nor do they represent the necessary logical order between them. It should also be understood that in the embodiments of the present application, "multiple" may refer to two or more, and "at least one" may refer to one, two or more. It should also be understood that for any component, data or structure mentioned in the embodiments of the present application, in the absence of explicit limitation or contrary revelation given in the context, it can generally be understood as one or more. In addition, the term "and / or" in the present application is only a kind of association relationship describing the associated objects, indicating that there can be three kinds of relationships, for example, A and / or B, which can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present application generally indicates that the associated objects before and after are an "or" relationship. It should also be understood that the description of each embodiment in the present application emphasizes the differences between the embodiments, and the same or similar parts can refer to each other. For the sake of brevity, they will not be repeated one by one.

[0019] At the same time, it should be understood that, for ease of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present application and its application or use. The techniques, methods and devices known to ordinary technicians in the relevant fields may not be discussed in detail, but where appropriate, the techniques, methods and devices should be considered as part of the specification. It should be noted that similar numbers and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

[0020] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0021] Figure 1A schematic flow chart of a method for preparing an oxidized low-density lipoprotein control product based on a novel antigen provided in an embodiment of the present application. It should be noted that in the process of preparing oxLDL quality control products, optimizing antigen preparation conditions and screening for optimal antigen epitopes through molecular dynamics simulation and machine learning algorithms are key innovations for improving the stability and uniformity of quality control products. This process mainly includes three steps: constructing an oxidized low-density lipoprotein model and determining the optimal antigen preparation conditions, predicting and screening for optimal antigen epitopes, and preparing quality control products based on a target cross-linking strategy.

[0022] like Figure 1 As shown, at step S101, an oxidized low-density lipoprotein model is constructed, and molecular dynamics simulation is performed by setting multiple sets of simulation conditions to calculate the stability index and determine the optimal antigen preparation conditions. Figure 2 ) which includes: constructing an oxidized low-density lipoprotein molecular model, including obtaining an initial structure based on low-density lipoprotein structural modification, and parameterizing the initial structure; setting multiple groups of simulation conditions, including pH value, temperature and ion concentration, and constructing multiple simulation systems based on the simulation conditions; performing energy minimization, equilibrium and production simulation on the multiple simulation systems, respectively, to obtain multiple groups of simulation trajectories; analyzing the multiple groups of simulation trajectories, calculating the root mean square deviation and the number of hydrogen bonds; calculating the stability index based on the root mean square deviation, the number of hydrogen bonds and the energy fluctuation; comparing the stability index under different simulation conditions, and selecting the condition with the highest stability index as the optimal antigen preparation condition.

[0023] First of all, constructing an oxidized low-density lipoprotein model and determining the optimal antigen preparation conditions through molecular dynamics simulation are the basis of the whole process. Oxidized low-density lipoprotein (oxLDL) is a complex biological molecule whose structure and properties are affected by many factors. By constructing an accurate molecular model and performing simulations, we can gain a deep understanding of the behavior of oxLDL under different conditions and optimize its preparation conditions. This step first requires the construction of a molecular model of oxLDL, which is usually based on the known low-density lipoprotein (LDL) structure and appropriately modified to simulate the oxidation process. The initial structure constructed needs to be parameterized to ensure that the model can accurately reflect the physicochemical properties of oxLDL. Parameterization may involve the adjustment of force field parameters, optimization of charge distribution, etc., which all require fine-tuning based on experimental data and theoretical calculations.

[0024] After the model is built, multiple sets of simulation conditions need to be set to explore the behavior of oxLDL in different environments. These conditions mainly include pH, temperature and ion concentration, which are all key factors affecting protein stability. pH affects the charge distribution of the protein, which in turn affects its structure and function; temperature is directly related to the thermal motion and conformational changes of the molecule; ion concentration affects electrostatic interactions and solvation effects. By systematically changing these parameters, a series of simulation systems can be constructed to comprehensively evaluate the stability of oxLDL.

[0025] For each simulation system, three stages of calculations are required: energy minimization, equilibrium, and production simulation. Energy minimization aims to eliminate unreasonable conformations that may exist in the initial structure, the equilibrium stage allows the system to reach a thermodynamic equilibrium state, and the production simulation is a long-term kinetic simulation based on the equilibrium state to obtain sufficient conformational sampling.

[0026] After the simulation is completed, the obtained trajectory needs to be analyzed and key stability indicators calculated. The root mean square deviation (RMSD) is an important parameter for evaluating the stability of protein structure, which reflects the degree of change of protein conformation over time. The number of hydrogen bonds is an indicator of the strength of internal interactions within the protein. The stability of the hydrogen bond network is directly related to the stability of the overall protein structure. In addition, energy fluctuations are also an important consideration. Smaller energy fluctuations usually mean a more stable structure.

[0027] Based on these indicators, a comprehensive stability index can be calculated. This index may be a weighted average of multiple parameters, and the weights may need to be adjusted based on experimental data or expert experience. By comparing the stability index under different simulation conditions, the optimal antigen preparation conditions can be determined. These conditions will provide important guidance for subsequent experimental preparation and help improve the stability of oxLDL quality control products.

[0028] At step S102, antigen epitope data is obtained and a feature vector is constructed to predict and screen the best antigen epitope using a machine learning algorithm. Figure 3 ) includes: obtaining known oxidized low-density lipoprotein antigen epitope data from an immune epitope database, including amino acid sequences and structural features thereof, using bioinformatics tools to process the sequence data and calculate secondary structure information; constructing a feature vector representing the known epitope, including amino acid composition, secondary structure composition, hydrophobicity index, isoelectric point, molecular weight and surface accessibility; based on the constructed feature vector, using an ensemble learning method to train a machine learning model to predict the epitope, and sorting and screening the prediction results; and preparing corresponding synthetic peptides or recombinant proteins according to the screened candidate epitopes by solid phase synthesis or recombinant protein expression.

[0029] The second step is to use machine learning algorithms to predict and screen the best antigen epitopes. Antigen epitopes are regions on antigen molecules that can be specifically recognized and bound by antibodies, and are essential for developing high-quality quality control products. Traditional methods for predicting antigen epitopes often rely on experimental screening, which is time-consuming, labor-intensive and inefficient. Machine learning methods can quickly and efficiently predict potential antigen epitopes, greatly accelerating the screening process.

[0030] This step first requires obtaining known oxLDL epitope data from the immune epitope database. These data usually include amino acid sequences and their structural features. The acquired data needs to be processed by bioinformatics tools to calculate secondary structure information, etc. This information provides the basis for the subsequent feature vector construction. The construction of feature vectors is the key to machine learning model training, and it needs to contain sufficient information to accurately describe the characteristics of antigen epitopes. Common features include amino acid composition, secondary structure composition, hydrophobicity index, isoelectric point, molecular weight and surface accessibility. These features describe the physicochemical properties and structural characteristics of antigen epitopes from different angles.

[0031] Based on the constructed feature vector, the machine learning model can be trained using ensemble learning methods. Ensemble learning improves prediction performance by combining multiple base learners. Commonly used methods include random forests and gradient boosting trees. These methods can effectively process high-dimensional features and are highly robust to noisy data. The trained model can be used to predict antigen epitopes for new sequences and sort and screen the prediction results.

[0032] The candidate epitopes screened need to be experimentally verified. Common methods include solid phase synthesis and recombinant protein expression. Solid phase synthesis is suitable for the preparation of short peptides, while recombinant protein expression is suitable for longer peptides or complete proteins. These experimental methods can quickly obtain actual samples of predicted epitopes, providing a material basis for the subsequent preparation of quality control products.

[0033] In step S103, based on the target cross-linking strategy, the oxidized low-density lipoprotein prepared under the optimal antigen preparation conditions is combined with the optimal antigen epitope to obtain the target quality control product. Figure 4 ) which includes: preparing oxidized low-density lipoprotein under preset pH, temperature and ionic strength conditions according to the results of the optimal antigen preparation conditions; preparing antigen epitope peptides by solid phase synthesis based on the optimal antigen epitope predicted by machine learning, and verifying the peptides by high performance liquid chromatography purification and mass spectrometry analysis; combining the prepared oxidized low-density lipoprotein with the synthesized antigen epitope peptides through a target chemical cross-linking reaction to produce a quality control product.

[0034] The last step is to combine the oxLDL prepared under optimal conditions with the optimal antigen epitope based on the target cross-linking strategy to prepare the target quality control product. This step first requires the preparation of oxLDL under specific pH, temperature and ionic strength conditions according to the optimal antigen preparation conditions determined previously. This process requires strict control of experimental conditions to ensure that the prepared oxLDL has good stability and uniformity. At the same time, based on the optimal antigen epitope predicted by machine learning, the corresponding peptide segment needs to be prepared by solid phase synthesis. Solid phase synthesis is an efficient peptide preparation method that can quickly obtain high-purity target peptide segments. The synthesized peptide segment needs to be purified by high-performance liquid chromatography to remove impurities generated during the synthesis process. The purified peptide segment also needs to be verified by mass spectrometry to ensure that its sequence and molecular weight are consistent with expectations.

[0035] Finally, the prepared oxLDL is combined with the synthesized epitope peptide through a targeted chemical cross-linking reaction. Chemical cross-linking is the process of connecting two or more molecules through covalent bonds, which can significantly improve the stability and uniformity of the quality control product. Common cross-linking strategies include the use of cross-linking agents such as glutaraldehyde, EDC / NHS, etc. The selection of an appropriate cross-linking strategy requires consideration of the chemical properties of oxLDL and the peptide, as well as the effect of the cross-linking reaction on the epitope structure.

[0036] Through the optimization of these three steps, the stability and uniformity of oxLDL quality control products can be significantly improved. Molecular dynamics simulation provides a theoretical basis for the optimization of antigen preparation conditions, machine learning algorithms improve the efficiency and accuracy of antigen epitope prediction, and the target cross-linking strategy ensures the high quality of the final quality control products.

[0037] Figure 2 A schematic diagram of a method for optimizing antigen preparation conditions using molecular dynamics simulation provided in an embodiment of the present application. Optimizing antigen preparation conditions using molecular dynamics simulation. This process can use GROMACS software (version 2021.4) to construct and simulate the oxLDL molecular model.

[0038] like Figure 2As shown, at step S201, an oxidized low-density lipoprotein molecular model is constructed, including obtaining an initial structure based on the modification of the low-density lipoprotein structure, and parameterizing the initial structure. Based on the LDL structure in the PDB database (PDB ID: 1N7D) as the starting model, the structure is modified using PyMOL software (version 2.5.2), and an oxidative modification group (such as 4-hydroxynonenal HNE) is added to the lysine residue to construct the oxLDL initial structure. Next, the CHARMM36 force field is used to parameterize the oxLDL molecular model. In order to more accurately describe the behavior of the oxidative modification group, it is necessary to use the quantum chemical calculation software Gaussian 16 to perform charge fitting on HNE to obtain the RESP charge. Then the antechamber module of AmberTools20 is used to convert the RESP charge into a CHARMM-compatible format.

[0039] At step S202, multiple sets of simulation conditions are set, including pH value, temperature and ion concentration, and multiple simulation systems are constructed based on the simulation conditions. For example, different simulation conditions are set: pH value range 5.0-8.0 (step length 0.5), temperature range 4-37°C (step length 5°C), NaCl concentration range 0-500 mM (step length 50 mM). For each combination of conditions, a simulation system including oxLDL molecules, water molecules and ions is constructed.

[0040] At step S203, the multiple simulation systems are respectively subjected to energy minimization, equilibrium and production simulation to obtain multiple sets of simulation trajectories. Specifically, each simulation system is subjected to energy minimization (5000 steps of steepest descent method, 5000 steps of conjugate gradient method), followed by 100 ps NVT equilibrium and 100 ps NPT equilibrium, and finally 100 ns of production simulation. The simulation adopts periodic boundary conditions, uses the Particle Mesh Ewald method to handle long-range electrostatic interactions, and the cutoff radius is set to 1.2 nm.

[0041] At step S204, the plurality of simulation trajectories are analyzed to calculate the root mean square deviation and the number of hydrogen bonds. For example, the simulation trajectories are analyzed using the analysis tools gmx rms and gmx hbond provided by GROMACS to calculate the RMSD and the number of hydrogen bonds. In order to quantify the structural stability, a stability index is introduced.

[0042] In step S205, a stability index is calculated based on the root mean square deviation, the number of hydrogen bonds and the energy fluctuation. Specifically, the stability index is calculated, including: ,in, represents the stability index, represents the root mean square deviation, represents the average number of hydrogen bonds, represents the theoretical maximum number of hydrogen bonds, represents the energy fluctuation of the system, represents the Boltzmann constant, Indicates absolute temperature.

[0043] At step S206, the stability indexes under different simulation conditions are compared, and the condition with the highest stability index is selected as the optimal antigen preparation condition. The SI under different conditions is compared, and the condition with the highest SI is selected as the optimal antigen preparation condition. For example, after simulation analysis, one of the optimal conditions is pH 7.2, 25°C, 150 mM NaCl.

[0044] Figure 3 A schematic flow diagram of a method for predicting and screening the best antigenic epitope using a machine learning algorithm provided in the present application embodiment. Next, the machine learning algorithm is applied to predict and screen the best antigenic epitope. This process can be achieved using Python 3.9 and the scikit-learn library (version 0.24.2).

[0045] like Figure 3 As shown, at step S301, known oxidized low-density lipoprotein epitope data, including amino acid sequence and structural features thereof, are obtained from the immune epitope database, and the sequence data are processed using bioinformatics tools and secondary structure information is calculated. For example, known oxLDL epitope data, including amino acid sequence, structural features (such as hydrophobicity, charge, secondary structure tendency, etc.) are collected from the immune epitope database IEDB. Sequence data are processed using the Biopython library (version 1.79), and secondary structure information is calculated using the DSSP program (version 3.0.0).

[0046] At step S302, a feature vector representing the known epitope is constructed, including amino acid composition, secondary structure composition, hydrophobicity index, isoelectric point, molecular weight and surface accessibility. Specifically, a feature vector is constructed. For each known epitope, the following features are calculated: a) Amino acid composition (20-dimensional vector) b) Secondary structure composition (3D vector, representing the proportion of helix, fold and random coil) c) Hydrophobicity index (using the Kyte-Doolittle scale) d) Isoelectric point e) Molecular weight f) Surface accessibility (calculated using DSSP) At step S303, based on the constructed feature vector, an ensemble learning method is used to train the machine learning model to predict the epitope, and the prediction results are sorted and screened. Wherein, based on the constructed feature vector, an ensemble learning method is used to train the machine learning model to predict the epitope, and the prediction results are sorted and screened, including: based on the constructed feature vector, an ensemble learning method is used to train the machine learning model, and grid search and cross-validation are used to optimize the hyperparameters of the model; a sliding window scan is performed on the complete sequence of oxidized low-density lipoprotein, the feature vector of each fragment is calculated, and the probability of it being an epitope is predicted by the trained machine learning model; an epitope scoring function is constructed, the predicted epitope probability is combined with the relative solvent accessible area and the shortest distance from the center of the fragment to the protein surface, and the weight coefficient of the function is optimized by a genetic algorithm; and the predicted epitopes are sorted and screened according to the constructed epitope scoring function.

[0047] In one embodiment, these features are used to train a machine learning model. Considering that the data set may be unbalanced, an integrated learning method Random Forest Classifier is used. Hyperparameters (such as the number of trees, maximum depth, etc.) are optimized using grid search and 5-fold cross validation. The complete sequence of oxLDL is scanned with a sliding window (window size is 15 amino acids, step size is 1), the above features are calculated for each fragment, and the trained model is used for prediction. In order to further improve the accuracy of the prediction, the epitope score function ES is introduced: ES = w1 * P(epitope) + w2 * (1 - SASA) + w3 * exp(-d / d0), where P(epitope) is the epitope probability predicted by the random forest model, SASA is the relative solvent accessible area (0-1 normalized), d is the shortest distance from the fragment center to the protein surface, and d0 is the characteristic distance (set to 5 Å). The weights w1, w2, and w3 are optimized by a genetic algorithm, and the objective function is to maximize the ES score of the known epitope.

[0048] Then, the predicted epitopes were ranked according to the ES scores, and the top five epitopes with the highest scores were selected as candidates. Structural analysis of these candidate epitopes was performed, and their positions and exposure in the three-dimensional structure of oxLDL could be visualized using PyMOL software.

[0049] At step S304, the corresponding synthetic peptide or recombinant protein is prepared by solid phase synthesis or recombinant protein expression according to the selected candidate epitope. Based on the selected epitope, the corresponding synthetic peptide or recombinant protein expression strategy is designed. For example, for synthetic peptides, the Fmoc solid phase synthesis method is used to synthesize on an ABI 433A peptide synthesizer. For recombinant proteins, the gene sequence containing the target epitope is designed, the codon usage is optimized, and it is cloned into a pET-28a vector and expressed in Escherichia coli BL21 (DE3).

[0050] Figure 4 A schematic diagram of a process for preparing a quality control product provided in an embodiment of the present application, that is, applying the results of computer-aided design to the actual antigen preparation process.

[0051] like Figure 4 As shown, at step S401, according to the result of the optimal antigen preparation conditions, oxidized low-density lipoprotein is prepared under preset pH, temperature and ionic strength conditions. In combination with the above embodiment, oxLDL is prepared according to the simulated optimized conditions (pH7.2, 25°C, 150 mM NaCl). The specific steps are as follows: LDL is isolated from the plasma of healthy volunteers using ultracentrifugation (density 1.019-1.063 g / mL). The separated LDL is dialyzed into PBS buffer (pH7.2) containing 150 mM NaCl at 4°C overnight. 5 μM CuSO4 is added and incubated at 37°C for 24 hours for oxidation. Excess Cu2+ is removed using a PD-10 desalting column (GE Healthcare). Protein concentration is determined using a BCA protein quantification kit (Thermo Scientific).

[0052] At step S402, based on the best antigenic epitope predicted by machine learning, an antigenic epitope peptide is prepared by solid phase synthesis, and the peptide is verified by high performance liquid chromatography purification and mass spectrometry analysis. For synthesizing or expressing the best antigenic epitope predicted by machine learning. Taking synthetic peptides as an example, the specific steps are as follows: Use Fmoc chemistry to synthesize the predicted epitope peptide on an ABI 433A peptide synthesizer. Use reverse phase HPLC (Waters 2695 system, C18 column) to purify the synthesized peptide. Use MALDI-TOF mass spectrometry (Bruker Autoflex III) to verify the molecular weight of the peptide.

[0053] At step S403, the prepared oxidized low-density lipoprotein is combined with the synthesized antigen epitope peptide segment through a target chemical cross-linking reaction to prepare a quality control product. The prepared oxLDL is combined with the synthesized antigen epitope to prepare a quality control product. In order to improve the binding efficiency and stability, a new cross-linking strategy is introduced: wherein the prepared oxidized low-density lipoprotein is combined with the synthesized antigen epitope peptide segment through a target chemical cross-linking reaction, including: ,in, stands for N-hydroxysuccinimide activated oxidized low-density lipoprotein, Indicates the epitope peptide with cysteine ​​added to the C-terminus. Represents a cross-linked product, Denotes N-hydroxysuccinimide, a by-product.

[0054] The details are as follows: Sulfo-NHS and EDC were used to activate the carboxyl groups on oxLDL (100 mM MES buffer, pH 6.0, room temperature for 30 minutes). The excess activator was removed using a PD-10 column. The activated oxLDL was reacted with the epitope peptide (molar ratio 1:20) in PBS buffer (pH 7.4) at 4°C overnight. Unbound peptides were removed using an ultrafiltration tube with a molecular weight cutoff of 100 kDa (Millipore).

[0055] In addition, after the quality control product is prepared, the method further includes: verifying the specificity of the quality control product by enzyme-linked immunosorbent assay, and evaluating its stability by introducing a stability factor; wherein the stability factor includes: ,in, represents the stability factor, represents the ELISA signal intensity after storage for t days, represents the initial ELISA signal intensity, Represents the attenuation constant in units of , Indicates storage time, Indicates the difference in denaturation temperature before and after storage. represents the initial denaturation temperature.

[0056] In one embodiment, the specificity and stability of the quality control product are verified by ELISA method. The specific steps are as follows: The prepared quality control product is coated on a 96-well plate with carbonate buffer (pH 9.6) at 4°C overnight. Block with PBST containing 1% BSA for 1 hour. Add different concentrations of anti-oxLDL monoclonal antibodies (such as 4E6) and incubate at room temperature for 2 hours. Add HRP-labeled secondary antibody and incubate at room temperature for 1 hour. Add TMB substrate for color development, add stop solution after 15 minutes, and measure OD value at 450 nm.

[0057] Figure 5 A schematic diagram of a molecular dynamics simulation result provided in an embodiment of the present application. The figure is a three-dimensional scatter plot showing the change in the stability index (SI) of the oxidized low-density lipoprotein (oxLDL) antigen under different pH values, temperatures and sodium chloride concentrations. Each scatter point in the figure represents a specific set of preparation conditions, and the color of the scatter point represents the SI value under the conditions. The brighter the color, the higher the SI value, that is, the better the antigen stability. The x-axis of the figure represents the pH value, ranging from 5.0 to 8.0; the y-axis represents the temperature, ranging from 4°C to 37°C; the z-axis represents the sodium chloride concentration, ranging from 0 to 500 mM. The entire graph presents a complex three-dimensional distribution, reflecting the combined effects of these three parameters on the stability of the oxLDL antigen.

[0058] It can be observed from the figure that the SI value shows a nonlinear trend with the change of conditions. In terms of pH value, the SI value reaches a higher level in the neutral to alkaline region (pH 7.0-7.5), which may be because the charge distribution of oxLDL molecules is more conducive to maintaining its structural stability in this pH range. The effect of temperature on SI is also obvious. In the range of 20-30℃, the SI value is generally high. This temperature range may be the best balance between the thermal motion and structural stability of oxLDL molecules. As for the sodium chloride concentration, the medium concentration (about 100-300 mM) seems to be more conducive to maintaining a high SI value, which may be because the appropriate ionic strength helps to shield the electrostatic repulsion between molecules without excessively interfering with the hydrogen bond network inside the molecule.

[0059] The red asterisk in the figure indicates the point with the highest SI value, which is the optimal antigen preparation condition. According to the legend, the conditions corresponding to this point should be pH about 7.2, temperature around 25°C, and sodium chloride concentration around 150 mM. This shows that under conditions close to physiological conditions, oxLDL antigen may have the best stability, which is of great significance for the development of stable and reliable quality control products.

[0060] From the overall distribution, the change of SI value in the figure shows a certain continuity and smoothness, which reflects the reliability of the molecular dynamics simulation results. The change of SI value does not show abrupt jumps, but shows a gradual trend, which is in line with the general law of physical and chemical processes. At the same time, some local fluctuations and ups and downs can also be observed, which may reflect the complexity of the conformational changes of oxLDL molecules under different conditions.

[0061] It is worth noting that the distribution of SI values ​​in the figure is not a simple unimodal distribution, but presents multiple local high-value areas. This complex distribution pattern suggests that the stability of oxLDL molecules is affected by multiple factors and there may be multiple metastable states. For example, under low temperature and high salt conditions, a relatively high SI value area also appeared, which may represent a metastable conformation different from the optimal conditions. This information is of great value for understanding the structural diversity and dynamic equilibrium of oxLDL molecules.

[0062] In summary, this computer-aided design method can significantly improve the preparation efficiency and quality of oxLDL quality control products. This method not only optimizes the antigen preparation conditions, but also predicts the optimal antigen epitope through machine learning algorithms, providing a more reliable standard for clinical diagnosis. The innovation of this method lies in the combination of molecular dynamics simulation, machine learning and experimental technology, realizing the optimization of the whole process from molecular level to macro preparation.

[0063] Furthermore, an embodiment of the present application also provides a device for preparing an oxidized low-density lipoprotein control product based on a novel antigen, comprising: a processor, a memory, and a system bus; the processor and the memory are connected via the system bus; the memory is used to store one or more programs, and the one or more programs include instructions, which, when executed by the processor, enable the processor to execute any of the above methods.

[0064] Furthermore, an embodiment of the present application also provides a computer program product, which, when executed on a terminal device, enables the terminal device to execute any of the above methods.

[0065] It can be known from the description of the above implementation mode that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product can be stored in a storage medium such as ROM / RAM, a disk, an optical disk, etc., including several instructions for enabling a computer device (which can be a personal computer, a server, or a network communication device such as a media gateway, etc.) to execute the methods described in the various embodiments of the present application or certain parts of the embodiments.

[0066] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0067] It should also be noted that, in the embodiments of the present application, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements includes not only those elements, but also includes other elements that are not clearly listed, or also includes elements inherent to such process, method, article or equipment. In the absence of more restrictions, the elements limited by the statement "comprise one..." do not exclude the presence of other identical elements in the process, method, article or equipment including the elements.

[0068] The above description of the disclosed embodiments enables professionals and technicians in the field to implement or use the present application. Various modifications to these embodiments will be apparent to professionals and technicians in the field, and the general principles defined in the embodiments of the present application can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown in the embodiments of the present application, but will conform to the widest range consistent with the principles and novel features disclosed in the embodiments of the present application.

Claims

1. A method for preparing an oxidized low-density lipoprotein control product based on a novel antigen, characterized in that: include: Construct an oxidized low-density lipoprotein model, perform molecular dynamics simulations by setting multiple sets of simulation conditions, calculate the stability index, and determine the optimal antigen preparation conditions; Obtain antigen epitope data and construct feature vectors to predict and screen the best antigen epitopes using machine learning algorithms; Based on the targeted cross-linking strategy, oxidized low-density lipoprotein prepared under optimal antigen preparation conditions was combined with the optimal antigen epitope to obtain the target quality control product.

2. The preparation method according to claim 1, characterized in that: in, The oxidized low-density lipoprotein model was constructed, and molecular dynamics simulations were performed by setting multiple sets of simulation conditions to calculate the stability index and determine the optimal antigen preparation conditions, including: Constructing an oxidized low-density lipoprotein molecular model, including obtaining an initial structure based on low-density lipoprotein structural modification, and performing parameterization processing on the initial structure; Setting multiple sets of simulation conditions, including pH value, temperature and ion concentration, and constructing multiple simulation systems based on the simulation conditions; Performing energy minimization, balance and production simulation on the multiple simulation systems respectively to obtain multiple groups of simulation trajectories; Analyzing the plurality of simulation trajectories, and calculating the root mean square deviation and the number of hydrogen bonds; calculating a stability index based on the RMS deviation, number of hydrogen bonds, and energy fluctuation; The stability indexes under different simulation conditions were compared, and the condition with the highest stability index was selected as the optimal antigen preparation condition.

3. The preparation method according to claim 1, characterized in that: in, Obtain antigen epitope data and construct feature vectors to predict and screen the best antigen epitopes using machine learning algorithms, including: Acquire known oxidized low-density lipoprotein antigen epitope data from an immune epitope database, including amino acid sequences and structural features thereof, and use bioinformatics tools to process the sequence data and calculate secondary structure information; Construct feature vectors representing known epitopes, including amino acid composition, secondary structure composition, hydrophobicity index, isoelectric point, molecular weight, and surface accessibility; Based on the constructed feature vectors, an ensemble learning method is used to train a machine learning model to predict epitopes, and the prediction results are sorted and screened; According to the screened candidate epitopes, the corresponding synthetic peptides or recombinant proteins are prepared by solid phase synthesis or recombinant protein expression.

4. The preparation method according to claim 1, characterized in that: in, Based on the targeted cross-linking strategy, the oxidized low-density lipoprotein prepared under the optimal antigen preparation conditions is combined with the optimal antigen epitope to obtain the target quality control product, including: According to the results of the optimal antigen preparation conditions, oxidized low-density lipoprotein was prepared under the conditions of preset pH value, temperature and ionic strength; Based on the optimal antigen epitope predicted by machine learning, the antigen epitope peptide was prepared by solid phase synthesis, and the peptide was verified by high performance liquid chromatography purification and mass spectrometry analysis; The prepared oxidized low-density lipoprotein is combined with the synthetic antigen epitope peptide fragment through a targeted chemical cross-linking reaction to prepare a quality control product.

5. The preparation method according to claim 2, characterized in that: in, Calculate stability indices including: ， in, represents the stability index, represents the root mean square deviation, represents the average number of hydrogen bonds, represents the theoretical maximum number of hydrogen bonds, represents the energy fluctuation of the system, represents the Boltzmann constant, Indicates absolute temperature.

6. The preparation method according to claim 3, characterized in that: in, Based on the constructed feature vector, an ensemble learning method is used to train a machine learning model to predict epitopes, and the prediction results are sorted and screened, including: Based on the constructed feature vectors, an ensemble learning method is used to train a machine learning model, and grid search and cross-validation are used to optimize the hyperparameters of the model; The complete sequence of oxidized low-density lipoprotein was scanned in a sliding window, the feature vector of each fragment was calculated, and the probability of it being an epitope was predicted using the trained machine learning model; constructing an epitope scoring function, combining the predicted epitope probability with the relative solvent accessible area and the shortest distance from the fragment center to the protein surface, and optimizing the weight coefficient of the function using a genetic algorithm; The predicted epitopes were ranked and screened according to the constructed epitope scoring function.

7. The preparation method according to claim 4, characterized in that: in, The prepared oxidized low-density lipoprotein is combined with the synthetic antigen epitope peptide fragments through a targeted chemical cross-linking reaction, including: ， in, stands for N-hydroxysuccinimide activated oxidized low-density lipoprotein, Indicates the epitope peptide with cysteine ​​added to the C-terminus. Represents a cross-linked product, Denotes N-hydroxysuccinimide, a by-product.

8. The preparation method according to claim 4, characterized in that: in, After the quality control product is prepared, it also includes: The specificity of the quality control product is verified by enzyme-linked immunosorbent assay, and its stability is evaluated by introducing a stability factor; wherein the stability factor includes: ， in, represents the stability factor, represents the ELISA signal intensity after storage for t days, represents the initial ELISA signal intensity, Represents the attenuation constant in units of , Indicates storage time, Indicates the difference in denaturation temperature before and after storage. represents the initial denaturation temperature.

Images