Method for detecting red blood cell folate metabolites by liquid chromatography mass spectrometry

Abstract

The application relates to the technical field of liquid chromatography-mass spectrometry, and particularly discloses a method for detecting red blood cell folate metabolites by using liquid chromatography-mass spectrometry, which can monitor the liquid chromatogram of supernatant to be injected in real time, and extract the chromatographic separation state by using an image analysis algorithm based on artificial intelligence and deep learning, so as to dynamically adjust the gradient parameters and reduce the need for manual intervention, thereby helping to optimize the separation process of target substances in a complex matrix. In particular, the method based on deep learning can automatically identify and quantify key features in the chromatogram, and intelligently dynamically adjust the gradient elution parameters according to the information, thereby facilitating more optimized substance separation and detection. For example, if it is found that the separation of a certain component is not ideal, the separation effect can be improved by adjusting the proportion of mobile phase B, thereby optimizing the gradient elution process.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of liquid chromatography mass spectrometry, and more particularly, to a method for detecting red blood cell folate metabolites by liquid chromatography mass spectrometry. BACKGROUND

[0002] Folic acid, as an essential B vitamin for the human body, its metabolic state is closely related to a variety of diseases and pregnancy health. Red blood cell folate level is an important indicator for assessing folate reserves in the body, but folate in red blood cells is mostly in the form of polyglutamic acid, which needs to be converted into monoglutamic acid by hydrolysis to be accurately detected. In the prior art, high-performance liquid chromatography tandem mass spectrometry (HPLC-MS / MS) has become the mainstream method for detecting folic acid and its metabolites due to its high sensitivity and specificity.

[0003] Chinese patent CN116223693A provides a method for determining folic acid and its metabolites in red blood cells by high-performance liquid chromatography tandem mass spectrometry, which can use HPLC-MS / MS analysis method to determine folic acid and its metabolites in red blood cells. However, the HPLC-MS / MS analysis method in the above patent adopts a fixed gradient elution program, which is difficult to adapt to the dynamic changes of interfering components in different sample matrices, resulting in unstable separation efficiency, redundant analysis time and other problems. Specifically, by presetting gradient parameters based on experience, it can realize simultaneous detection of multiple components, but when facing complex biological samples, fixed gradient may lead to insufficient separation of target and interfering substances, and repeated optimization of conditions is required. In addition, existing dynamic gradient adjustment techniques mostly rely on manual experience or simple feedback mechanisms, lack intelligent optimization capabilities based on real-time data, and are difficult to adaptively adjust the proportion of mobile phase in a single run, limiting the analysis throughput and accuracy.

[0004] Therefore, an optimized liquid chromatography mass spectrometry method for detecting red blood cell folate metabolites is desired. SUMMARY

[0005] The present application provides a method for detecting red blood cell folate metabolites by liquid chromatography mass spectrometry, which can automatically identify and quantify key features in the chromatogram, and intelligently dynamically adjust the gradient elution parameters according to these information, thereby facilitating more optimized substance separation and detection.

[0006] According to one aspect of the present application, a method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry is provided, comprising: collecting a blood sample containing an EDTA anticoagulant, and performing centrifugal separation processing to remove the plasma to obtain a sample to be tested; pretreating the sample to be tested to obtain supernatant for injection; performing liquid chromatography analysis on the supernatant for injection to determine the proportion of the mobile phase; performing gradient elution on the supernatant for injection based on the proportion of the mobile phase to separate and remove interfering groups in the sample matrix to obtain a target analyte solution; detecting the mass-to-charge ratio of the target analyte and its corresponding isotopic internal standard in the target analyte solution by mass spectrometer, and using isotopic internal standard method for quantitative determination to determine the content of red blood cell folate metabolites; wherein the liquid chromatography analysis of the supernatant for injection to determine the proportion of the mobile phase comprises: obtaining a liquid chromatogram of the supernatant for injection; pretreating and liquid chromatography multi-dimensional feature extraction are performed on the liquid chromatogram to obtain liquid chromatogram shallow features and liquid chromatogram semantic features; liquid chromatography multi-dimensional feature joint coding is performed on the liquid chromatogram shallow features and the liquid chromatogram semantic features to obtain significant information interaction representation between liquid chromatogram shallow-semantic features, including: core clue extraction is performed on the liquid chromatogram shallow features and the liquid chromatogram semantic features respectively to obtain liquid chromatogram semantic core clue coding features and liquid chromatogram shallow core clue coding features; based on the liquid chromatogram semantic core clue coding features and the liquid chromatogram shallow core clue coding features, fine-grained feature interaction analysis is performed on the liquid chromatogram shallow features and the liquid chromatogram semantic features to obtain significant information interaction representation between the liquid chromatogram shallow-semantic features; and the proportion of the mobile phase is determined based on the significant information interaction representation between the liquid chromatogram shallow-semantic features.

[0007] In the above method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry, the pretreatment and liquid chromatography multi-dimensional feature extraction of the liquid chromatogram to obtain liquid chromatogram shallow features and liquid chromatogram semantic features comprises: denoising and baseline correction are performed on the liquid chromatogram to obtain a corrected liquid chromatogram; the corrected liquid chromatogram is input into a liquid chromatography feature extractor based on a cavity space pyramid pooling module to obtain a liquid chromatogram shallow feature map as the liquid chromatogram shallow features and a liquid chromatogram semantic feature map; and the liquid chromatogram semantic feature map is input into a liquid chromatogram semantic global perception reinforcement module based on a fully connected layer to obtain a liquid chromatogram semantic feature vector as the liquid chromatogram semantic features.

[0008] In the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology, the liquid chromatogram shallow feature and the liquid chromatogram semantic feature are respectively subjected to core clue extraction to obtain liquid chromatogram semantic core clue coding features and liquid chromatogram shallow core clue coding features, including: performing point convolution coding-based core clue extraction on the liquid chromatogram semantic feature vector to obtain a liquid chromatogram semantic core clue coding vector as the liquid chromatogram semantic core clue coding feature; and extracting a liquid chromatogram shallow core clue coding vector from the liquid chromatogram shallow feature map as the liquid chromatogram shallow core clue coding feature.

[0009] In the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology, the liquid chromatogram shallow feature and the liquid chromatogram semantic feature are respectively subjected to core clue extraction to obtain liquid chromatogram semantic core clue coding features and liquid chromatogram shallow core clue coding features, including: performing point convolution coding-based core clue extraction on the liquid chromatogram semantic feature vector to obtain a liquid chromatogram semantic core clue coding vector as the liquid chromatogram semantic core clue coding feature; and extracting a liquid chromatogram shallow core clue coding vector from the liquid chromatogram shallow feature map as the liquid chromatogram shallow core clue coding feature.

[0010] In the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology, the liquid chromatogram shallow feature and the liquid chromatogram semantic feature are respectively subjected to core clue extraction to obtain liquid chromatogram semantic core clue coding features and liquid chromatogram shallow core clue coding features, including: performing point convolution coding-based core clue extraction on the liquid chromatogram semantic feature vector to obtain a liquid chromatogram semantic core clue coding vector as the liquid chromatogram semantic core clue coding feature; and extracting a liquid chromatogram shallow core clue coding vector from the liquid chromatogram shallow feature map as the liquid chromatogram shallow core clue coding feature.

[0011] In the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology, based on the liquid chromatogram semantic core clue coding vector and the liquid chromatogram shallow core clue coding vector, a liquid chromatogram shallow-semantic inter-feature core clue weaving template matrix is constructed, including: constructing an inter-modal core clue weaving template between the liquid chromatogram semantic core clue coding vector and the liquid chromatogram shallow core clue coding vector to obtain the liquid chromatogram shallow-semantic inter-feature core clue weaving template matrix.

[0012] In the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology, the inter-modal core clue weaving template between the liquid chromatogram semantic core clue coding vector and the liquid chromatogram shallow core clue coding vector is constructed to obtain the liquid chromatogram shallow-semantic inter-feature core clue weaving template matrix, including: using a feature mapping function to process the liquid chromatogram semantic core clue coding vector and the liquid chromatogram shallow core clue coding vector to obtain a liquid chromatogram shallow-semantic inter-feature core association matrix; and performing semantic alignment compensation on the liquid chromatogram shallow-semantic inter-feature core association matrix to obtain the liquid chromatogram shallow-semantic inter-feature core clue weaving template matrix.

[0013] In the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology, the semantic alignment compensation is performed on the liquid chromatogram shallow-semantic inter-feature core association matrix to obtain the liquid chromatogram shallow-semantic inter-feature core clue weaving template matrix, including: using a residual mechanism and the feature mapping function to perform structured semantic alignment on the liquid chromatogram semantic core clue coding vector and the liquid chromatogram shallow core clue coding vector to obtain a liquid chromatogram semantic alignment vector and a liquid chromatogram shallow alignment vector; calculating an explicit correlation matrix between the liquid chromatogram semantic alignment vector and the liquid chromatogram shallow alignment vector; and based on the liquid chromatogram shallow-semantic inter-feature explicit correlation matrix, performing significant compensation on the liquid chromatogram shallow-semantic inter-feature core association matrix to obtain the liquid chromatogram shallow-semantic inter-feature core clue weaving template matrix.

[0014] The application provides a method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology, which can monitor the liquid chromatogram of the supernatant to be injected in real time, and extract the chromatographic separation state by using an image analysis algorithm based on artificial intelligence and deep learning, so as to dynamically adjust the gradient parameters and reduce the need for manual intervention, thereby helping to optimize the separation process of target substances in a complex matrix. In particular, the deep learning-based method can automatically identify and quantify key features in the chromatogram, and intelligently dynamically adjust the gradient elution parameters according to these information, thereby facilitating more optimized substance separation and detection. For example, if it is found that the separation of a certain component is not ideal, the separation effect of the component can be improved by adjusting the proportion of mobile phase B, thereby optimizing the gradient elution process. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings of the embodiments of the application will be briefly introduced below. Obviously, the drawings described below only relate to some embodiments of the application, but not limit the application.

[0016] Figure 1 A schematic flow chart of the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology of the embodiments of the application.

[0017] Figure 2 A schematic flow chart of step S3 in the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology of the embodiments of the application.

[0018] Figure 3 A schematic flow chart of step S32 in the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology of the embodiments of the application.

[0019] Figure 4 A schematic flow chart of step S33 in the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology of the embodiments of the application.

[0020] Figure 5 A schematic flow chart of step S331 in the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology of the embodiments of the application.

[0021] Figure 6 A schematic flow chart of step S332 in the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry technology of the embodiments of the application. DETAILED DESCRIPTION

[0022] With reference to the drawings, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort belong to the scope of the present application.

[0023] With the increasing application of machine learning algorithms in analytical chemistry, it is possible to optimize chromatographic separation processes using these advanced computing tools. However, existing research has focused on offline data modeling, and has not achieved online monitoring and optimization of HPLC-MS / MS systems, and has insufficient joint analysis capability for multi-dimensional chromatographic characteristics (such as peak shape, retention time drift, and co-elution interference), which limits its application in actual detection.

[0024] Therefore, in the technical solutions of the present application, a method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry is proposed, as shown in Figure 1 As shown in the figure, the method for detecting red blood cell folate metabolites by liquid chromatography-mass spectrometry comprises the following steps: S1, collecting a blood sample containing an EDTA anticoagulant and centrifuging and separating the blood sample to remove plasma to obtain a sample to be tested; S2, pretreating the sample to be tested to obtain supernatant for sampling; S3, performing liquid chromatography analysis on the supernatant for sampling to determine the proportion of the mobile phase; S4, gradient elution of the supernatant for sampling based on the proportion of the mobile phase to separate and remove interfering groups in the sample matrix to obtain a target analyte solution; and S5, detecting the mass-to-charge ratio of the target analyte and its corresponding isotopic internal standard in the target analyte solution by mass spectrometer, and using isotopic internal standard method for quantitative determination to determine the content of red blood cell folate metabolites.

[0025] For example, in step S1, a blood sample containing an EDTA anticoagulant is collected and centrifuged and separated to remove plasma to obtain a sample to be tested. It should be understood that the blood sample containing an EDTA anticoagulant is selected to prevent the blood from coagulating during collection and processing. EDTA is an effective chelating agent that can bind calcium ions, thereby preventing blood clotting factors from functioning, ensuring that the blood remains in a flowing state, facilitating subsequent processing. Once the blood sample containing the EDTA anticoagulant is collected, the next step is to remove the plasma by centrifugation to obtain a red blood cell sample. This process is necessary because it is necessary to specifically detect folate and its metabolites in red blood cells. Red blood cells are the main storage site of folate in the human body, and they contain a large amount of folate polyglutamate form, which must be converted to monoglutamate form to be accurately detected. Therefore, in order to ensure the accuracy of the analysis results, it is necessary to separate the red blood cells from the whole blood and further process them.

[0026] In one specific example of the present application, the collected blood sample is placed in a high-speed centrifuge and centrifuged under specific speed and time conditions. For example, the commonly used conditions can be 3000 to 5000 revolutions per minute (rpm) for 10 to 20 minutes. This centrifugation process causes the blood components to stratify according to their density differences: the lighter plasma is at the top, and the heavier red blood cells are deposited at the bottom. By carefully removing the upper layer of plasma, a pure red blood cell sample can be obtained. It is worth noting that special care should be taken when removing the plasma to avoid mixing in any residual plasma or other components such as white blood cells, which can affect the accuracy of subsequent experiments.

[0027] Illustratively, in step S2, the sample to be tested is pretreated to obtain the supernatant for injection. It should be understood that because folate in red blood cells mainly exists in the form of polyglutamate, these forms need to be converted into monoglutamate form to be accurately detected. In addition, the red blood cell sample may contain various proteins and other impurities, which will interfere with subsequent chromatographic and mass spectrometric analysis. Therefore, by pretreatment, these interfering substances can be effectively removed, and folate and its metabolites can be converted into a form suitable for detection.

[0028] In one specific example of the present application, after obtaining a pure red blood cell sample, a mixed internal standard working solution is first added. This internal standard solution contains isotopically labeled compounds similar in structure but different in mass to the target analytes, such as folate-d4, 5-methyltetrahydrofolate-d4, etc. The addition of internal standards helps to correct errors that may occur during sample preparation and analysis, thereby improving the accuracy of quantitative results. Next, in order to protect folate and its metabolites from oxidative degradation, a protective agent, usually 1% to 3% ascorbic acid (VC) solution, is also added. At the same time, in order to adjust the pH value to be suitable for the subsequent enzymatic reaction, an appropriate amount of sodium hydroxide (NaOH) solution is added. These reagents work together to provide a stable chemical environment for the subsequent steps. Subsequently, in order to convert folate polyglutamate into monoglutamate, the red blood cell sample must be treated with a cleavage enzyme. The specific method is to add an aqueous solution containing 0.3 to 1.0 units / milliliter of glutamyl hydrolase and vortex mix well. Then, the mixture is incubated at 37°C in the dark for a period of time (such as 1 hour) to allow sufficient hydrolysis of folate polyglutamate. After completing the enzymatic reaction, in order to precipitate the proteins and further purify the sample, a protein precipitant is added. The commonly used precipitant is a methanol-acetonitrile mixed solution with a volume ratio of 2:1. After mixing by shaking, high-speed centrifugation (such as 14800 revolutions per minute for 5 minutes) is performed. The supernatant after centrifugation is the desired injection liquid, which contains folate and its metabolites converted into monoglutamate form and removes most of the proteins and other impurities.

[0029] Exemplarily, in step S3, the supernatant to be injected is subjected to liquid chromatography analysis to determine the proportion of mobile phase. It should be understood that the optimization of the proportion of mobile phase is crucial for achieving efficient chromatographic separation. Different compounds exhibit different retention behaviors under different mobile phase conditions, so selecting the appropriate proportion of mobile phase can significantly improve the separation effect of the target compound. Specifically, by adjusting the proportion of mobile phase A (usually an aqueous solution) and mobile phase B (usually an organic solvent such as methanol or acetonitrile), the retention time and elution order of the target compound in the chromatographic column can be controlled. This is particularly important for the simultaneous detection of multiple target substances in complex matrices, as different target substances may have different polarities and chemical properties, requiring specific mobile phase proportions to achieve optimal separation. Secondly, liquid chromatography analysis helps to identify and quantify key features in the chromatogram, providing a basis for dynamically adjusting gradient parameters. By real-time monitoring and analysis of the liquid chromatogram of the supernatant to be injected, multi-dimensional information about peak shape, retention time drift, co-elution interference, etc. can be obtained. These information can help researchers understand the separation effect under the current separation conditions and make adjustments according to the actual situation. For example, if the separation of a certain component is found to be unsatisfactory, the proportion of mobile phase B can be adjusted to improve its separation effect.

[0030] Exemplarily, in step S4, the supernatant to be injected is subjected to gradient elution based on the proportion of the mobile phase to separate the target analyte solution from the interfering components in the sample matrix. It should be understood that because the red blood cell sample contains a large amount of other components in addition to the target compound (such as folic acid and its metabolites), these components may interfere with the detection of the target compound. For example, proteins, lipids, and other small molecule metabolites may co-elute with the target compound, causing signal overlap or increased background noise, thereby affecting the detection results. By gradient elution, the composition of the mobile phase can be gradually changed, so that compounds of different polarity are eluted at different time points, thereby achieving effective separation. This method is particularly suitable for complex matrix samples, as it can dynamically adjust the mobile phase conditions according to the characteristics of the target compound to optimize the separation effect.

[0031] Exemplarily, in step S5, the mass-to-charge ratio of the target analyte and its corresponding isotopic internal standard in the target analyte solution is detected by a mass spectrometer, and the isotopic internal standard method is used for quantification to determine the content of red blood cell folate metabolites. It should be understood that the mass spectrometer can provide extremely high sensitivity and specificity, which is crucial for the detection of low-concentration target compounds in complex biological matrices. In particular, in red blood cell samples, the concentration of folate and its metabolites is low, and there are a large number of interfering substances. The mass spectrometer can achieve accurate identification and quantification of target compounds by accurately measuring the mass-to-charge ratio (m / z). The mass spectrometer usually uses the multiple reaction monitoring (MRM) mode for positive ion scanning. In this mode, the instrument selects specific precursor ions and product ions for detection, greatly improving the selectivity and sensitivity of the detection. This method is particularly suitable for simultaneous detection of multiple compounds such as folate, 5-methyltetrahydrofolate, 5-formyltetrahydrofolate, etc. In order to reduce the errors introduced during sample preparation and analysis, isotopically labeled internal standards are usually added. These internal standards have similar chemical properties to the target analytes, but have different mass-to-charge ratios when detected by mass spectrometry. By comparing the response intensity of the target analyte and the internal standard, errors caused by sample handling, injection volume changes, etc. can be corrected, thereby improving the accuracy of quantification.

[0032] In one specific example of the present application, first, the target analyte solution after high-performance liquid chromatography (HPLC) separation enters the mass spectrometer. The mass spectrometer ionizes the compounds in the solution by electrospray ionization (ESI) mode. In this process, the compound molecules are ionized into charged particles, and then separated and detected in the mass spectrometer according to their mass-to-charge ratio (m / z). Specifically, for each target compound and its corresponding isotopic internal standard, the mass spectrometer will scan within a specific mass-to-charge ratio range and record their signal intensities. For example, in this method, the target compounds of folate and its metabolites include folic acid (VB9), 5-methyltetrahydrofolate (5-MTHF), 5-formyltetrahydrofolate (5-FoTHF), flavin adenine dinucleotide (FAD), and vitamin B3 (VB3). The corresponding isotopic internal standards are folic acid-d4 (VB9-d4), 5-methyltetrahydrofolate-d4 (5-MTHF-d4), 5-formyltetrahydrofolate-d4 (5-FoTHF-d4), flavin adenine dinucleotide-d5 (FAD-d5), and vitamin B3-d4 (VB3-d4), respectively. During the mass spectrometer detection process, the capillary voltage is set to 3.0 kV, the nozzle voltage is 200 V, the dry gas temperature is 300°C, the sheath gas temperature is 350°C, the dry gas flow rate is 10 L / min, the sheath gas flow rate is 11 L / min, and the nebulizer gas pressure is 45 psi. Optimization of these parameters helps improve ionization efficiency and detection sensitivity. Next, the mass spectrometer records the mass-to-charge ratio response of the target compounds and their isotopic internal standards. By comparing the response intensities of the target compounds and the internal standards, the concentration of the target compounds can be calculated. Specifically, according to the pre-established standard curve, the concentration of the target compound is calculated using the internal standard method. The standard curve is obtained by adding known concentrations of standard and internal standard to the blank matrix, measuring their response intensities, and plotting them. In this way, even if there are some minor changes in the sample processing process, the internal standard can be used for correction to ensure the accuracy of the quantitative results. Finally, by repeatedly detecting and processing data from multiple samples, the accurate content of folic acid and its metabolites in red blood cells can be obtained. This method is not only suitable for laboratory research, but also can be used for clinical diagnosis and health assessment, helping doctors better understand the nutritional status and metabolic conditions of patients.

[0033] Specifically, in the above method, the process of analyzing the supernatant to be injected for liquid chromatography to determine the proportion of the mobile phase is crucial, but the gradient elution program in the existing scheme is fixed, which limits the separation and detection process of the target substance in the face of different sample characteristics. Based on this, the technical concept of the present application is to be able to monitor the liquid chromatogram of the supernatant to be injected in real time, and use image analysis algorithms based on artificial intelligence and deep learning to extract the chromatographic separation state, thereby dynamically adjusting the gradient parameters and reducing the need for manual intervention, thereby helping to optimize the separation process of the target substance in a complex matrix. In particular, the deep learning-based method can automatically identify and quantify key features in the chromatogram, and intelligently dynamically adjust the gradient elution parameters based on this information, thereby facilitating more optimized substance separation and detection. For example, if it is found that the separation of a certain component is not ideal, the separation effect can be improved by adjusting the proportion of mobile phase B, thereby optimizing the gradient elution process.

[0034] In one embodiment, as shown in FIG. 3, in step S3, analyzing the supernatant to be injected for liquid chromatography to determine the proportion of the mobile phase comprises: Figure 2 S31, obtaining a liquid chromatogram of the supernatant to be injected; S32, preprocessing and liquid chromatography multi-dimensional feature extraction are performed on the liquid chromatogram to obtain liquid chromatogram shallow features and liquid chromatogram semantic features; S33, liquid chromatography multi-dimensional feature joint coding is performed on the liquid chromatogram shallow features and the liquid chromatogram semantic features to obtain significant information interaction representation between liquid chromatogram shallow-semantic features; S34, determining the proportion of the mobile phase based on the significant information interaction representation between the liquid chromatogram shallow-semantic features.

[0035] Exemplarily, in step S31, a liquid chromatogram of the supernatant to be injected is obtained. It should be understood that the liquid chromatogram of the supernatant to be injected is obtained to ensure that the target compound can be effectively separated and detected in a complex matrix. In the red blood cell sample, in addition to the target folate and its metabolites, there are a large number of other components such as proteins, lipids and other small molecule metabolites. These interfering substances can co-elute with the target compound, causing signal overlap or increased background noise, thereby affecting the accuracy and sensitivity of the detection results. By obtaining the liquid chromatogram, the separation of each component can be observed directly, potential interference peaks can be identified, and corresponding measures can be taken for optimization. Secondly, the liquid chromatogram can provide important information about the components of the sample. For example, by analyzing the peak shape, retention time, peak area and other parameters in the chromatogram, the separation effect of different compounds and the presence of co-elution can be evaluated. This is particularly important for simultaneous detection of multiple components in a complex matrix, because different target compounds may have different polarity and chemical properties, and require specific mobile phase ratios to achieve optimal separation. In addition, the liquid chromatogram can also be used to correct experimental conditions to ensure consistency and reliability of each analysis. Finally, the liquid chromatogram can be analyzed in combination with artificial intelligence and deep learning algorithms to further improve the separation effect. Modern computing tools can automatically identify and quantify key features in the chromatogram, and intelligently adjust the gradient elution parameters based on this information. This method not only captures significant interaction information between shallow features (such as peak shape, retention time, etc.) and semantic features (such as the chemical properties of target compounds, metabolic pathways, etc.) in the chromatogram, but also provides more comprehensive data support for subsequent mobile phase ratio optimization. In this way, the mobile phase ratio can be adaptively adjusted in a single run, thereby achieving more optimized substance separation and detection.

[0036] In one specific example of the present application, as shown in Figure 3 In step S32, the liquid chromatogram is preprocessed and liquid chromatography multi-dimensional feature extraction is performed to obtain liquid chromatogram shallow features and liquid chromatogram semantic features, including: S321, denoising and baseline correction are performed on the liquid chromatogram to obtain a corrected liquid chromatogram; S322, the corrected liquid chromatogram is passed through a liquid chromatography feature extractor based on a hollow space pyramid pooling module to obtain a liquid chromatogram shallow feature map as the liquid chromatogram shallow feature and a liquid chromatogram semantic feature map; S323, the liquid chromatogram semantic feature map is passed through a liquid chromatogram semantic global perception reinforcement module based on a fully connected layer to obtain a liquid chromatogram semantic feature vector as the liquid chromatogram semantic feature.

[0037] Exemplarily, in step S321, it is considered that in liquid chromatography analysis, the liquid chromatogram is an important data source for detecting target compounds. However, due to factors such as instrument noise, baseline drift, background interference, etc., the original liquid chromatogram often contains a large amount of noise and irregular baseline fluctuation, which seriously affects the subsequent data analysis and result accuracy. Therefore, the liquid chromatogram is denoised and baseline corrected to obtain a corrected liquid chromatogram. Specifically, the noise in the liquid chromatogram can be derived from instrument electronic noise, detector fluctuation, flow phase inhomogeneity, etc. The noise can mask the signal of the target compound, and the signal of the low-concentration target can be overwhelmed by the noise. Through denoising processing, the noise level can be effectively reduced, the peak signal of the target compound can be highlighted, the signal-to-noise ratio can be improved, and the target can be more accurately identified and quantified. In addition, baseline drift can be caused by factors such as flow phase gradient change, column temperature fluctuation, detector response change, etc. Baseline drift can cause peak distortion, affecting the accurate judgment of peak integral area and retention time. Through baseline correction, the baseline can be restored to a stable state, ensuring the accurate starting point and ending point of each peak, avoiding peak area calculation errors caused by baseline drift, and improving the accuracy of quantitative analysis.

[0038] In a specific example of the present application, first, the original liquid chromatogram containing noise and baseline drift is obtained. These noises can be derived from instrument electronic noise, detector fluctuation, flow phase inhomogeneity, etc., while baseline drift can be caused by factors such as flow phase gradient change, column temperature fluctuation, or detector response change. In order to improve the signal-to-noise ratio and ensure the accuracy of the peak shape, these interference factors need to be processed. In order to remove background noise, a smoothing filter method is used. Specifically, after loading the original liquid chromatogram in the data analysis software (such as Agilent OpenLab CDS or Thermo Scientific Chromeleon), a suitable smoothing algorithm is selected, such as the Savitzky-Golay filter. This filter can reduce the influence of high-frequency noise while preserving the main characteristics of the signal. By setting an appropriate window width (such as 5 points or 7 points), the noise level can be effectively reduced, and the peak signal of the target compound can be clearer. Another method is to use wavelet transform technology. This method can decompose the signal at different scales, effectively distinguishing noise and useful signals. By selecting a suitable wavelet type (such as db4) and decomposition level (such as 4 layers), the liquid chromatogram can be decomposed. Then, thresholding processing is applied to each decomposition coefficient to remove noise components below a certain threshold. Finally, the processed coefficients are reconstructed into a denoised liquid chromatogram. The chromatogram processed in this way not only reduces noise, but also maintains the main characteristics of the target peak.

[0039] After denoising, baseline correction is needed next. Baseline drift can cause peak shape distortion, affecting the accurate judgment of peak integral area and retention time. Therefore, it is necessary to restore the baseline to a stable state. Common baseline correction methods include manual correction and automatic correction. In manual correction, by observing the baseline trend in the chromatogram, a relatively flat area can be selected as the reference baseline, and then the baseline of other parts is adjusted to be consistent with the reference baseline. This method is suitable for simple chromatograms, but for complex multi-component samples, manual correction may consume a lot of time and effort. In contrast, automatic baseline correction is more efficient. Modern data analysis software usually has built-in automatic baseline correction functions that can intelligently identify and correct the baseline according to the characteristics of the chromatogram. For example, in Agilent OpenLab CDS, you can select the "automatic baseline correction" option, and the software will automatically identify the baseline drift and adjust it to a stable state. In addition, some parameters can be set to optimize the correction effect, such as baseline threshold, smoothing window size, etc.

[0040] Exemplarily, in step S322, after the liquid chromatogram is corrected, in order to capture the characteristics of the target substance in the complex matrix in the liquid chromatogram and monitor the separation in real time to dynamically adjust the proportion of the mobile phase, feature extraction needs to be performed on the corrected liquid chromatogram. However, in the traditional convolutional neural network, the receptive field size is fixed, which limits the model's understanding of the context information on a larger range of liquid chromatograms. Therefore, in the technical solution of the present application, the corrected liquid chromatogram is further processed by a liquid chromatogram feature extractor based on a cavity space pyramid pooling module to obtain a liquid chromatogram shallow feature map and a liquid chromatogram semantic feature map. Through the processing of the liquid chromatogram feature extractor based on the cavity space pyramid pooling module, the network can capture liquid chromatogram features at different scales, which is particularly important for complex and variable liquid chromatogram data. This is because the peak shape, retention time drift, co-elution interference and other information in the liquid chromatogram may exhibit different characteristics at different scales, which helps to enhance the model's understanding ability of different sample characteristics, thereby better achieving control of the gradient elution process. It is worth mentioning that the liquid chromatogram shallow feature map usually contains more original detail information about the input liquid chromatogram, such as the position, width and preliminary morphological characteristics of the peak, which is very important for identifying specific compound peaks. The liquid chromatogram semantic feature map focuses more on high-level abstract features, such as the relationship between different compounds, overall distribution patterns, etc., which are very important for understanding the composition and dynamic changes of complex mixtures.

[0041] Exemplarily, in step S323, since the liquid chromatogram semantic feature map contains multi-scale information about the liquid chromatogram semantic features, the different scales of the liquid chromatogram semantics are related to the characteristics perception and subsequent gradient elution control of different samples, but these information can be relatively scattered, and it is difficult to present the global representation information about the semantic features in the liquid chromatogram semantic feature map. Based on this, in the technical solution of the present application, the liquid chromatogram semantic feature map is further processed by a liquid chromatogram semantic global perception reinforcement module based on a full connection layer to obtain a liquid chromatogram semantic feature vector. Through the processing of the liquid chromatogram semantic global perception reinforcement module based on the full connection layer, the complex relationship and semantic correlation features between different regions in the liquid chromatogram can be captured, which is particularly important for understanding complex liquid chromatograms. Such more global liquid chromatogram semantic feature representation is crucial for distinguishing different compounds and their metabolites.

[0042] Exemplarily, in step S33, the liquid chromatogram shallow features and the liquid chromatogram semantic features are jointly encoded in a liquid chromatogram multi-dimensional feature to obtain significant information interaction representation between the liquid chromatogram shallow features and the liquid chromatogram semantic features. It should be understood that in liquid chromatography analysis, the liquid chromatogram is the core data source for detecting target compounds. In order to extract more rich and accurate feature information from the chromatogram, not only the chromatogram needs to be analyzed and encoded from multiple dimensions (such as shallow features and semantic features), but also the significant interaction information between the shallow features (such as peak shape, retention time, etc.) and the semantic features (such as the chemical properties of the target compound, the metabolic pathway, etc.) in the chromatogram needs to be captured, so as to provide more accurate guidance for subsequent mobile phase proportion optimization. Specifically, the shallow features of the liquid chromatogram mainly include intuitive information such as peak retention time, peak height, peak area, peak shape, etc. These features can reflect the physical and chemical properties (such as polarity, molecular weight, etc.) of the target compound and the separation state (such as whether it is overlapped with other peaks). The semantic features of the liquid chromatogram are a deep understanding of the chromatogram, such as the chemical structure of the target compound, the metabolic pathway, the interaction with other compounds, etc. Based on this, in the technical solution of the present application, the liquid chromatogram shallow features and the liquid chromatogram semantic features are further jointly encoded in a liquid chromatogram multi-dimensional feature to obtain significant information interaction representation between the liquid chromatogram shallow features and the liquid chromatogram semantic features. Through the processing of the liquid chromatogram multi-dimensional feature joint encoding, the significant interaction relationship between the shallow features and the semantic features in the liquid chromatogram can be captured, such as whether the separation effect of a certain peak is affected by the chemical structure of the target compound, or whether a certain baseline drift is related to the gradient change of the mobile phase. These interaction relationships can provide more comprehensive data support for subsequent mobile phase proportion optimization, thereby helping to optimize the gradient elution program and dynamically adjust the mobile phase proportion. For example, if it is found that the separation effect of a certain peak is not ideal, the model can adjust the proportion of mobile phase B according to the information in the interaction encoding vector, thereby helping to improve the separation effect.

[0043] In one embodiment, as shown in FIG. 33, Figure 4 In one embodiment, as shown in FIG. 33,

[0044] In one specific example of the present application, as shown in Figure 5 In step S331, the liquid chromatogram shallow feature and the liquid chromatogram semantic feature are respectively subjected to core clue extraction to obtain liquid chromatogram semantic core clue coding features and liquid chromatogram shallow core clue coding features, including: S3311, performing point convolution coding-based core clue extraction on the liquid chromatogram semantic feature vector to obtain a liquid chromatogram semantic core clue coding vector as the liquid chromatogram semantic core clue coding feature; S3312, extracting a liquid chromatogram shallow core clue coding vector from the liquid chromatogram shallow feature map as the liquid chromatogram shallow core clue coding feature.

[0045] Exemplarily, in step S3311, the process of performing point convolution coding-based core clue extraction on the liquid chromatogram semantic feature vector to obtain a liquid chromatogram semantic core clue coding vector aims to extract the most representative information segments from complex raw data. This process is like an experienced analyst searching for those most critical signals in a vast amount of data, which can most effectively describe the characteristics of the sample and the presence of each component therein. Specifically, directly using these liquid chromatogram semantic feature vectors for subsequent analysis may face many challenges, including but not limited to noise interference, complexity of co-elution peaks, and variability under different experimental conditions. In order to solve these problems, it is crucial to adopt the method of point convolution coding to extract core clues from semantic feature vectors. The point convolution operation here plays the role of a refiner, which not only maintains the integrity of high-dimensional information, but also strengthens those features that are semantically significant. This means that after point convolution coding, the originally complex chromatogram is simplified into a highly condensed and abstract representation - namely the semantic core clue coding vector. This representation makes subsequent analysis more focused on truly important features rather than being overwhelmed by irrelevant information. In addition, point convolution effectively avoids redundant expression of information, which achieves this by focusing on important elements of the modality (in this case, the chromatogram). In this process, point convolution can identify and emphasize those parts that are most critical for understanding the composition of the sample, such as the unique peak shape of a particular metabolite or the interaction pattern between it and adjacent components. Doing so not only improves the sensitivity of detection, but also enhances the specificity of results, thereby more accurately quantifying the concentration of target compounds. Specifically, the calculation process of the liquid chromatogram semantic core clue coding vector can be represented by the formula:

[0046]

[0047] wherein, is the liquid chromatogram semantic feature vector, It is point convolution processing. It is a bias vector. yes Activation function It is the semantic core clue encoding vector of liquid chromatogram.

[0048] In a specific example of this application, in step S3312, extracting the liquid chromatogram shallow core cue encoding vector from the liquid chromatogram shallow feature map as the liquid chromatogram shallow core cue encoding feature includes: splitting the liquid chromatogram shallow feature map to obtain a sequence of liquid chromatogram shallow local feature maps. Specifically, this process can be expressed by the formula:

[0049]

[0050] in, This is a shallow feature map of the liquid chromatogram. Indicates feature map splitting, These represent the first, second, and third segments of the sequence representing the shallow local feature maps of a liquid chromatogram. The and the first A shallow local feature map of a liquid chromatogram.

[0051] Attention-based salient region attention analysis is performed on each shallow local feature map in the sequence of liquid chromatograms to obtain a sequence of salient region attention coefficients. Specifically, this process can be expressed by the formula:

[0052]

[0053]

[0054]

[0055] in, This is a convolutional layer with a 3×3 kernel. for Activation function For the first A shallow, locally hidden feature map of a liquid chromatography chromatogram. To perform global pooling on each feature matrix along the channel dimension in the feature map. For the first The shallow local feature vector of a liquid chromatogram corresponds to the shallow local hidden feature map of the liquid chromatogram. and These represent the maximum and minimum values ​​in the vector, respectively. , and Both are weighting coefficients. and are the mean and variance of the vector, respectively, is the salient region attention focus coefficient.

[0056] Based on the sequence of salient region attention focus coefficients, a core clue search is performed on the liquid chromatogram shallow feature map to obtain a liquid chromatogram shallow core clue encoding vector. Specifically, the process can be represented by the formula:

[0057]

[0058] wherein, is the exponential function value with the natural constant e as the base, is the number of feature maps in the sequence of liquid chromatogram shallow local feature maps, is the liquid chromatogram shallow core clue encoding vector.

[0059] Exemplarily, in step S3312, the entire process of splitting, salient region attention analysis, and core clue search on the liquid chromatogram shallow feature map aims to extract the most critical information segments from complex and multi-dimensional data sets. This process is crucial for accurately identifying and quantifying folic acid and its metabolites in red blood cells. First, splitting the original liquid chromatogram shallow feature map is to better understand and process the data. Imagine if you directly process the entire chromatogram, it may be difficult to capture subtle but important details due to the overwhelming amount of information. By splitting, a complex image can be decomposed into multiple local feature maps, each focusing on a specific time period or component. This is like dividing a large painting into small blocks for careful study, each block can be observed and analyzed in more detail, making it easier to find valuable information hidden in a large amount of data.

[0060] Next, attention-based salient region attention analysis is applied to these split liquid chromatogram shallow local feature maps. Here, the attention mechanism is like a pair of automatic focusing eyes installed on the system, which can automatically identify the most important part among numerous information. In this process, the algorithm will evaluate each local feature map and give different weights according to its importance. For example, when analyzing red blood cell samples, those peak shapes related to target folic acid and its metabolites will be given higher attention because they are key indicators for judging sample quality. This method not only improves the efficiency of data analysis, but also ensures that the most relevant information is given priority, reducing noise and other irrelevant factors interference.

[0061] Finally, based on the sequence of significant region attention focus coefficients, the shallow feature map of the liquid chromatogram is searched for core clues, with the purpose of extracting the most representative feature vector, i.e., the liquid chromatogram shallow core clue encoding vector. This step is like the process of further refining gold after preliminary screening. By integrating the most significant parts of each local feature map, a highly condensed representation can be constructed, which contains the most important and most useful information about the sample.

[0062] In one specific example of the present application, as shown in Figure 6 In one specific example of the present application, as shown in

[0063] In one specific example of the present application, in step S3321, based on the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector, the liquid chromatogram shallow-semantic inter-feature core clue weaving template matrix is constructed, including: constructing the inter-modal core clue weaving template between the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector to obtain the liquid chromatogram shallow-semantic inter-feature core clue weaving template matrix.

[0064] In one specific example of the present application, constructing the inter-modal core clue weaving template between the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector to obtain the liquid chromatogram shallow-semantic inter-feature core clue weaving template matrix includes: using a feature mapping function to process the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector to obtain a liquid chromatogram shallow-semantic inter-feature core association matrix, specifically, the process can be represented by the formula:

[0065]

[0066] wherein, and are feature mapping functions, such as linear mapping or non-linear kernel functions, respectively, denotes transposition, is a core association matrix between shallow-semantic features of liquid chromatogram. The core association matrix between shallow-semantic features of liquid chromatogram is compensated by semantic alignment to obtain a core clue weaving template matrix between shallow-semantic features of liquid chromatogram.

[0067] In a specific example of the present application, the core association matrix between shallow-semantic features of liquid chromatogram is compensated by semantic alignment to obtain a core clue weaving template matrix between shallow-semantic features of liquid chromatogram, comprising: structuring semantic alignment of the semantic core clue coding vector of liquid chromatogram and the shallow core clue coding vector of liquid chromatogram by using a residual mechanism and the feature mapping function to obtain a semantic alignment vector of liquid chromatogram and a shallow alignment vector of liquid chromatogram, specifically, the process can be represented by formula as:

[0068]

[0069]

[0070] wherein, denotes positional subtraction, and denote the semantic alignment vector of liquid chromatogram and the shallow alignment vector of liquid chromatogram, respectively.

[0071] An explicit correlation matrix between shallow-semantic features of liquid chromatogram is calculated between the semantic alignment vector of liquid chromatogram and the shallow alignment vector of liquid chromatogram, specifically, the process can be represented by formula as:

[0072]

[0073] wherein, is the square of the two-norm of the calculation vector, denotes the logarithmic function value with base 2, is an explicit correlation matrix between shallow-semantic features of liquid chromatogram.

[0074] The core association matrix between shallow-semantic features of liquid chromatogram is significantly compensated based on the explicit correlation matrix between shallow-semantic features of liquid chromatogram to obtain the core clue weaving template matrix between shallow-semantic features of liquid chromatogram, specifically, the process can be represented by formula as:

[0075]

[0076] wherein, is a core clue weaving template matrix between the shallow-semantic features of the liquid chromatogram.

[0077] Exemplarily, in step S3321, it is necessary to first integrate the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector. This is like combining two different perspectives or two different sets of observations to form a more comprehensive view. For example, when analyzing a red blood cell sample, on the one hand there is confirmation information about the presence of folic acid and its metabolites (from semantic features), and on the other hand there is a detailed description of the specific forms of these compounds in the chromatogram (from shallow features). By constructing an inter-modal core clue weaving template, a bridge can be built between these two types of information, so that each type of information can complement and support the other. Next, the process of generating a core clue weaving template matrix using this weaving template is actually creating a framework that can effectively organize and integrate data from different sources. It is like a smart map or indexing system that helps quickly find the most useful parts of the vast amount of information and presents them in a structured way. This not only helps improve the efficiency of data analysis, but also enhances the reliability and accuracy of the results.

[0078] Preferably, in the above process, the feature vectors of the core clue representations for different modal features (such as semantic features and shallow features) and are used to perform preliminary structured semantic alignment of general explicit association in the shared representation space through their residual mechanism based on dissimilar modal mapping to obtain the liquid chromatogram semantic alignment vector and the liquid chromatogram shallow alignment vector and . Then, based on the explicit modeling graph structure global semantic spatialization representation under mapping preference, the residual mechanism is further used to capture the "missing" information in the alignment process under dissimilar mapping modalities and adaptively compensate through explicit association to further capture the inter-modal fine-grained saliency difference information in the preliminary alignment process. In this way, the contradiction between the alignment complexity of cross-modal dissimilar intrinsic driving mapping and the intuitive nature of explicit association of graph structure can be optimized, and the inter-modal global semantic constraint illusion under the condition of fine-grained relationship misalignment between local salient regions can be reduced.

[0079] In a specific example of the present application, in step S3322, based on the core clue interweaving template matrix between the shallow features and the semantic features of the liquid chromatogram, cross-modal fine-grained feature interaction analysis is performed on the semantic feature vector of the liquid chromatogram and the shallow feature map of the liquid chromatogram to obtain a significant information interaction encoding vector between the shallow features and the semantic features of the liquid chromatogram as the significant information interaction representation between the shallow features and the semantic features of the liquid chromatogram, including: performing feature decoupling along the channel dimension on the shallow feature map of the liquid chromatogram to obtain a set of local feature matrices of the liquid chromatogram, specifically, the process can be represented by the formula:

[0080]

[0081] wherein, represents feature decoupling along the channel dimension, respectively represent each local feature matrix of the sequence of local feature matrices of the liquid chromatogram.

[0082] The semantic feature vector of the liquid chromatogram is taken as a query vector, each local feature matrix in the set of local feature matrices of the liquid chromatogram is taken as a key matrix, and the core clue interweaving template matrix between the shallow features and the semantic features of the liquid chromatogram is taken as a prior information constraint matrix, which are input into the cross-modal template constraint encoder based on the heterogeneous converter to obtain a set of fine-grained interaction encoding vectors between the shallow features and the semantic features of the liquid chromatogram, specifically, the process can be represented by the formula:

[0083]

[0084] wherein, is the i-th local feature matrix of the sequence of local feature matrices of the liquid chromatogram, is the dimension of , i.e., the width multiplied by the height of the matrix, is matrix multiplication, represents function, is the i-th fine-grained interaction encoding vector in the set of fine-grained interaction encoding vectors between the shallow features and the semantic features of the liquid chromatogram. The position-wise mean vector of the set of fine-grained interaction encoding vectors between the shallow features and the semantic features of the liquid chromatogram is calculated to obtain the significant information interaction encoding vector between the shallow features and the semantic features of the liquid chromatogram, specifically, the process can be represented by the formula:

[0085]

[0086] ​​

[0087] wherein, is the number of vectors in the set of shallow-semantic feature interaction encoding vectors of the liquid chromatogram, is the significant information interaction encoding vector between the shallow features and the semantic features of the liquid chromatogram.

[0088] Exemplarily, in step S3322, the shallow feature map of the liquid chromatogram is first decoupled along the channel dimension to obtain a set of shallow local feature matrices of the liquid chromatogram. This process decomposes the complex original feature map into multiple local feature matrices. Each local feature matrix represents detailed information of the chromatogram in a specific time period or component, so that subsequent analysis can capture the key features of each local area in more detail. Then, taking the semantic feature vector of the liquid chromatogram as the query vector, each matrix in the set of shallow local feature matrices of the liquid chromatogram as the key matrix, and combining the shallow-semantic feature interaction core clue weaving template matrix as the prior information constraint matrix, input them into the cross-modal template constraint encoder based on the heterogeneous converter. In this process, the semantic feature vector acts as the role of query, helping the system to focus on the information needed; while the shallow local feature matrix provides specific detailed information. Through the cross-modal template constraint encoder of the heterogeneous converter, these information is integrated to form a set of a series of fine-grained interaction encoding vectors. This step not only ensures the effective interaction between different modalities, but also strengthens the normativity and accuracy of this interaction through the prior information constraint matrix, improving the representation ability of the results. Finally, the position-wise mean vector of the set of shallow-semantic feature interaction fine-grained encoding vectors of the liquid chromatogram is calculated to obtain the significant information interaction encoding vector between the shallow features and the semantic features of the liquid chromatogram. This process further compresses the processing cost and retains the main semantic information. Specifically, by calculating the position-wise mean of all fine-grained interaction encoding vectors, a highly condensed and information-rich representation form is generated - the significant information interaction encoding vector between the shallow features and the semantic features of the liquid chromatogram. This step not only simplifies the workload of subsequent analysis, but also improves the accuracy and reliability of the results.

[0089] Exemplarily, in step S34, the proportion of the mobile phase is determined based on the significant information interaction representation between the shallow-semantic features of the liquid chromatogram. In one specific example of the present application, determining the proportion of the mobile phase based on the significant information interaction representation between the shallow-semantic features of the liquid chromatogram includes: passing the significant information interaction encoding vector between the shallow-semantic features of the liquid chromatogram through a decoder-based mobile phase proportion generator to obtain a mobile phase proportion recommended decoding value. That is, the significant information interaction semantics between the shallow features and the semantic features of the liquid chromatogram are used for decoding regression to optimize the gradient elution program and adjust the proportion of the mobile phase. In this way, the gradient parameters can be dynamically adjusted, reducing the need for manual intervention, thereby helping to optimize the separation process of the target substance in a complex matrix and facilitating more optimized substance separation and detection.

[0090] In one specific example of the present application, in order to convert the significant information interaction encoding vector between the shallow-semantic features of the liquid chromatogram into a mobile phase proportion recommended decoding value, a specific full-connection decoder-based mobile phase proportion generator can be designed. The input layer of this decoder receives the significant information interaction encoding vector as input and finally outputs the recommended value of the proportion of the mobile phase through a layer of full-connection layer. Among them, the number of nodes of the output layer is determined according to the number of the proportion of the mobile phase to be predicted. For example, if the proportions of two solvents A and B are to be predicted, the output layer should have two nodes to represent the proportions of the two solvents respectively.

[0091] In one specific example of the present application, when performing reversed-phase high-performance liquid chromatography (RP-HPLC), water containing 0.1% trifluoroacetic acid is used as phase A and acetonitrile containing 0.1% trifluoroacetic acid is used as phase B. Under the initial setting, the proportion of the mobile phase can be 95:5 (A:B), meaning 95% water and 5% acetonitrile. If the experimental results show that some compounds cannot be well separated, the system will automatically adjust the proportion of the mobile phase according to the significant information interaction encoding vector. After calculation by the full-connection decoder-based mobile phase proportion generator, the new recommended value is 70:30 (A:B). This means that starting from the initial 95:5, the proportion will gradually change to 70:30 in the next 20 minutes to enhance the elution ability of the solvent, ensuring that those compounds that are difficult to dissolve in water can be effectively eluted. In this process, by intelligently adjusting the proportion of the mobile phase, not only can the separation conditions be optimized, the analysis efficiency and accuracy be improved, but also the need for manual intervention can be reduced, making the whole process more automated and intelligent.

[0092] In summary, the method for detecting red blood cell folate metabolites by liquid chromatography mass spectrometry technology according to the embodiments of the present application is illustrated, which can monitor the liquid chromatogram of the supernatant to be injected in real time, and extract the chromatographic separation state by using the image analysis algorithm based on artificial intelligence and deep learning, so as to dynamically adjust the gradient parameters and reduce the need for manual intervention, thereby helping to optimize the separation process of the target in the complex matrix. In particular, the method based on deep learning can automatically identify and quantify the key features in the chromatogram, and intelligently dynamically adjust the gradient elution parameters according to these information, thereby facilitating more optimized material separation and detection. For example, if it is found that the separation of a certain component is not ideal, the separation effect can be improved by adjusting the proportion of mobile phase B, thereby optimizing the gradient elution process.

[0093] Finally, it should be noted that the above-described embodiments are part of the embodiments of the present application, rather than all the embodiments. The detailed description of the embodiments of the present application is not intended to limit the scope of the claimed application, but only represents selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative labor belong to the scope of protection of the present application.

Claims

1. A method for detecting red blood cell folate metabolites by liquid chromatography mass spectrometry, comprising: Collect blood samples containing EDTA anticoagulant, and centrifuge the samples to remove plasma to obtain the test sample; The test sample is pretreated to obtain a supernatant sample; The supernatant sample is analyzed by liquid chromatography to determine the proportion of the mobile phase; based on the proportion of the mobile phase, the supernatant sample is gradient eluted to separate and remove the interfering groups in the sample matrix to obtain a target analyte solution; The mass-to-charge ratio of the target analyte in the target analyte solution and its corresponding isotopic internal standard is detected by a mass spectrometer, and the isotopic internal standard method is used for quantitative determination to determine the content of red blood cell folate metabolites; characterized in that the liquid chromatography analysis of the supernatant sample to determine the proportion of the mobile phase comprises: Obtain the liquid chromatogram of the supernatant sample; The liquid chromatogram is preprocessed and liquid chromatography multi-dimensional feature extraction is performed to obtain liquid chromatogram shallow features and liquid chromatogram semantic features, including: denoising and baseline correction of the liquid chromatogram to obtain a corrected liquid chromatogram; the corrected liquid chromatogram is passed through a liquid chromatography feature extractor based on a hollow space pyramid pooling module to obtain a liquid chromatogram shallow feature map as the liquid chromatogram shallow feature and a liquid chromatogram semantic feature map; the liquid chromatogram semantic feature map is passed through a liquid chromatogram semantic global perception reinforcement module based on a fully connected layer to obtain a liquid chromatogram semantic feature vector as the liquid chromatogram semantic feature; Liquid chromatography multi-dimensional feature joint coding is performed on the liquid chromatogram shallow features and the liquid chromatogram semantic features to obtain significant information interaction representation between liquid chromatogram shallow-semantic features, including: core clue extraction is performed on the liquid chromatogram shallow features and the liquid chromatogram semantic features respectively to obtain liquid chromatogram semantic core clue coding features and liquid chromatogram shallow core clue coding features, including: core clue extraction based on point convolution coding is performed on the liquid chromatogram semantic feature vector to obtain a liquid chromatogram semantic core clue coding vector as the liquid chromatogram semantic core clue coding feature; a liquid chromatogram shallow core clue coding vector is extracted from the liquid chromatogram shallow feature map as the liquid chromatogram shallow core clue coding feature; Based on the liquid chromatogram semantic core clue coding feature and the liquid chromatogram shallow core clue coding feature, fine-grained feature interaction analysis is performed on the liquid chromatogram shallow features and the liquid chromatogram semantic features to obtain significant information interaction representation between the liquid chromatogram shallow-semantic features; Based on the significant information interaction representation between the liquid chromatogram shallow-semantic features, the proportion of the mobile phase is determined.

2. The method for detecting red blood cell folate metabolites by liquid chromatography mass spectrometry technique according to claim 1, characterized in that, Extracting a liquid chromatogram shallow core clue coding vector from the liquid chromatogram shallow feature map as the liquid chromatogram shallow core clue coding feature comprises: Splitting the liquid chromatogram shallow feature map to obtain a sequence of liquid chromatogram shallow local feature maps; performing attention-based significant region attention analysis on each of the sequence of liquid chromatogram shallow local feature maps to obtain a sequence of significant region attention coefficients; performing core clue searching on the liquid chromatogram shallow feature map based on the sequence of significant region attention coefficients to obtain a liquid chromatogram shallow core clue encoding vector.

3. The method for detecting red blood cell folate metabolites by liquid chromatography mass spectrometry technique according to claim 2, characterized in that, performing fine-grained feature interaction analysis on the liquid chromatogram shallow feature and the liquid chromatogram semantic feature based on the liquid chromatogram semantic core clue encoding feature and the liquid chromatogram shallow core clue encoding feature to obtain significant information interaction representation between the liquid chromatogram shallow feature and the liquid chromatogram semantic feature, including: constructing a core clue interweaving template matrix between the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector based on the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector; performing cross-modal fine-grained feature interaction analysis on the liquid chromatogram semantic feature vector and the liquid chromatogram shallow feature map based on the liquid chromatogram shallow-semantic feature inter-core clue interweaving template matrix to obtain a significant information interaction encoding vector as the significant information interaction representation between the liquid chromatogram shallow feature and the liquid chromatogram semantic feature.

4. The method for detecting red blood cell folate metabolites by liquid chromatography mass spectrometry according to claim 3, wherein, constructing a core clue interweaving template matrix between the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector based on the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector, including: constructing an inter-modal core clue interweaving template between the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector to obtain the liquid chromatogram shallow-semantic feature inter-core clue interweaving template matrix.

5. The method for detecting red blood cell folate metabolites by liquid chromatography mass spectrometry according to claim 4, wherein, constructing a core clue interweaving template matrix between the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector based on the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector, including: processing the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector using a feature mapping function to obtain a liquid chromatogram shallow-semantic feature inter-core association matrix; performing semantic alignment compensation on the liquid chromatogram shallow-semantic feature inter-core association matrix to obtain the liquid chromatogram shallow-semantic feature inter-core clue interweaving template matrix.

6. The method for detecting red blood cell folate metabolites by liquid chromatography mass spectrometry according to claim 5, wherein, performing semantic alignment compensation on the liquid chromatogram shallow-semantic feature inter-core association matrix to obtain the liquid chromatogram shallow-semantic feature inter-core clue interweaving template matrix, including: performing structured semantic alignment on the liquid chromatogram semantic core clue encoding vector and the liquid chromatogram shallow core clue encoding vector using a residual mechanism and the feature mapping function to obtain a liquid chromatogram semantic alignment vector and a liquid chromatogram shallow alignment vector; calculating an explicit correlation matrix between the liquid chromatogram semantic alignment vector and the liquid chromatogram shallow alignment vector; Based on the explicit correlation matrix between the shallow-semantic features of the liquid chromatogram, the core correlation matrix between the shallow-semantic features of the liquid chromatogram is significantly compensated to obtain a core clue weaving template matrix between the shallow-semantic features of the liquid chromatogram.

7. The method for detecting red blood cell folate metabolites by liquid chromatography mass spectrometry according to claim 6, wherein, Based on the core clue weaving template matrix between the shallow-semantic features of the liquid chromatogram, cross-modal fine-grained feature interaction analysis is performed on the semantic feature vector of the liquid chromatogram and the shallow feature map of the liquid chromatogram to obtain a significant information interaction encoding vector between the shallow-semantic features of the liquid chromatogram as the significant information interaction representation between the shallow-semantic features of the liquid chromatogram, including: Performing feature decoupling along the channel dimension on the shallow feature map of the liquid chromatogram to obtain a set of liquid chromatogram shallow local feature matrices; Taking the semantic feature vector of the liquid chromatogram as a query vector, each liquid chromatogram shallow local feature matrix in the set of liquid chromatogram shallow local feature matrices as a key matrix, and the core clue weaving template matrix between the shallow-semantic features of the liquid chromatogram as a prior information constraint matrix, inputting it into a cross-modal template constraint encoder based on a heterogeneous converter to obtain a set of fine-grained interaction encoding vectors between the shallow-semantic features of the liquid chromatogram. Calculating the position-wise mean vector of the set of fine-grained interaction encoding vectors between the shallow-semantic features of the liquid chromatogram to obtain the significant information interaction encoding vector between the shallow-semantic features of the liquid chromatogram.

8. The method for detecting red blood cell folate metabolites by liquid chromatography mass spectrometry technique according to claim 7, characterized in that, Based on the significant information interaction representation between the shallow-semantic features of the liquid chromatogram, determining the flow phase ratio, including passing the significant information interaction encoding vector between the shallow-semantic features of the liquid chromatogram through a decoder-based flow phase ratio generator to obtain a flow phase ratio recommendation decoding value.

Images