Liquid chromatography-tandem mass spectrometry detection method and system

Abstract

The invention discloses a liquid chromatography-tandem mass spectrometry detection method and system, and the method comprises the steps: designing a multi-dimensional gradient elution program according to the physicochemical properties of a target compound, and obtaining a chromatographic separation spectrogram; performing efficient ionization on the target compound according to the chromatographic separation spectrogram to obtain an ionization signal; according to the ionization signal, fragmenting the target ions, and obtaining a fragment ion spectrogram through fragment ion screening; according to the fragment ion spectrogram, a multivariable data analysis algorithm based on machine learning is adopted, a standard curve of a target compound and an interferent database are combined, quantitative analysis is carried out, and a quantitative result is generated through a dynamic interference correction technology in combination with principal component analysis and a partial least square method. By utilizing the embodiment of the invention, the sensitivity, the accuracy and the real-time performance of detection can be improved through multi-dimensional gradient elution, dynamic ionization and fragmentation parameter optimization in combination with an analysis means of machine learning.

Description

Technical Field

[0001] The present invention belongs to the technical field of mass spectrometry detection, and particularly relates to a liquid chromatography tandem mass spectrometry detection method and system. Background Art

[0002] Liquid chromatography tandem mass spectrometry (LC-MS) technology, as an important analytical tool, is widely used in fields such as drug analysis, environmental monitoring, and food safety. Its core lies in the effective separation and detection of target compounds in samples. However, existing LC-MS methods still face various challenges. Firstly, traditional liquid chromatography separation often relies on fixed gradient elution programs, making it difficult to efficiently separate different target compounds in complex samples, resulting in insufficient reproducibility and accuracy of chromatographic separation spectra. In addition, the proportion and flow rate of the mobile phase usually lack real-time optimization and cannot meet the requirements of dynamic separation. Secondly, during the ionization process, the efficiency of electrospray ionization is affected by various factors. Traditional methods have not effectively improved the ionization sensitivity of low-concentration target compounds, which may lead to signal loss or interference. Although the fragmentation technology in mass spectrometry analysis can provide more accurate quantitative results, existing methods rely on experience to select fragmentation energy, lacking systematicness and reducing the reliability of analysis. Finally, quantitative methods based on traditional statistical analysis often struggle to handle complex data, especially in the presence of interfering substances, and the accuracy of quantitative results is significantly affected. Summary of the Invention

[0003] The purpose of the present invention is to provide a liquid chromatography tandem mass spectrometry detection method and system to address the deficiencies in the prior art, capable of improving the sensitivity, accuracy, and real-time performance of detection through multi-dimensional gradient elution, dynamic ionization, and fragmentation parameter optimization, as well as analytical means combining machine learning.

[0004] An embodiment of the present application provides a liquid chromatography tandem mass spectrometry detection method, and the method includes: Design a multi-dimensional gradient elution program according to the physical and chemical properties of the target compound, use the elution program to perform liquid chromatography separation on the sample, and in combination with a dynamic gradient optimization algorithm, adjust the proportion and flow rate of the mobile phase in real time to obtain a chromatographic separation spectrum; According to the chromatographic separation spectrum, adopt an ionization efficiency enhancement technology based on electrospray ionization, and in combination with a dynamic voltage tuning and solvent-assisted ionization strategy, perform efficient ionization on the target compound, and dynamically adjust the ion source parameters through an adaptive electric field optimization algorithm to obtain an ionization signal; According to the ionization signal, adopt a multi-stage mass spectrometry fragmentation technology, and in combination with a dynamic energy tuning algorithm, fragment the target ion, and through fragment ion screening, in combination with a fragmentation rule library of the target compound, dynamically select characteristic fragment ions to obtain a fragment ion spectrum; According to the fragment ion spectrum, a multivariate data analysis algorithm based on machine learning is adopted, combined with the standard curve of the target compound and the interferent database, for quantitative analysis. Through dynamic interference correction technology, combined with principal component analysis and partial least squares method, quantitative results are generated.

[0005] Optionally, according to the physical and chemical properties of the target compound, a multi-dimensional gradient elution program is designed, and the sample is separated by liquid chromatography using the elution program. Combining with the dynamic gradient optimization algorithm, the mobile phase ratio and flow rate are adjusted in real time to obtain a chromatographic separation spectrum, including: According to the physical and chemical properties of the target compound, a prediction model based on molecular descriptors is used to calculate its retention time and elution conditions, and a preliminary elution condition range is generated through a multi-objective optimization algorithm. For the preliminary elution condition range, a multi-dimensional gradient elution program generation algorithm is adopted, combined with the three-dimensional parameter space of time-mobile phase ratio-flow rate, a gradient elution program is designed, and a preliminary elution program is generated through dynamic constraint conditions. According to the preliminary elution program, the dynamic gradient optimization algorithm is adopted, combined with the real-time chromatographic peak shape monitoring data, the mobile phase ratio and flow rate are adjusted, and the separation effect is optimized through a feedback control mechanism to generate a preliminary high-resolution chromatographic separation spectrum. For the preliminary high-resolution chromatographic separation spectrum, a peak shape optimization algorithm is adopted, combined with baseline correction and noise filtering techniques, to further improve the spectrum quality, and a final high-resolution chromatographic separation spectrum is generated through a resolution evaluation index.

[0006] Optionally, according to the chromatographic separation spectrum, an ionization efficiency enhancement technology based on electrospray ionization is adopted, combined with dynamic voltage tuning and solvent-assisted ionization strategies, to perform efficient ionization on the target compound. Through an adaptive electric field optimization algorithm, the ion source parameters are dynamically adjusted to obtain an ionization signal, including: According to the high-resolution chromatographic separation spectrum, combined with the physical and chemical properties of the target compound, the electrospray ionization conditions are initialized, and preliminary ionization parameters are generated through an ionization efficiency prediction model. For the preliminary ionization parameters, a dynamic voltage tuning algorithm is adopted, combined with the solvent-assisted ionization strategy, to adjust the spray voltage and solvent ratio in real time, and preliminary optimized ionization conditions are generated through ionization efficiency monitoring technology. For the preliminary optimized ionization conditions, an adaptive electric field optimization algorithm is adopted, combined with ion signal intensity feedback, to dynamically adjust the ion source parameters, and preliminary high-quality ionization signals are generated through electric field distribution simulation technology. For the preliminary high-quality ionization signals, signal-to-noise ratio and ion intensity stability evaluation indexes are used to optimize the signal quality, and final high-quality ionization signals are generated through signal filtering technology.

[0007] Optionally, according to the ionization signal, a multi-stage mass spectrometry fragmentation technique is adopted, combined with a dynamic energy tuning algorithm, to fragment the target ions. Through fragment ion screening, combined with the fragmentation rule library of the target compound, characteristic fragment ions are dynamically selected to obtain a fragment ion spectrum, including: According to the high-quality ionization signal, combined with the molecular weight and structural characteristics of the target compound, the fragmentation energy is initialized, and preliminary fragmentation parameters are generated through a fragmentation efficiency prediction model; For the preliminary fragmentation parameters, a dynamic energy tuning algorithm is adopted, combined with fragment ion intensity feedback, to adjust the collision energy in real time. Through a multi-stage fragmentation strategy, a preliminary fragment ion spectrum is generated; For the preliminary fragment ion spectrum, a screening algorithm based on the fragmentation rule library is adopted, combined with the characteristic fragments of the target compound, to dynamically select characteristic fragment ions. Through fragment ion matching technology, a preliminary highly specific fragment ion spectrum is generated; For the preliminary highly specific fragment ion spectrum, signal-to-noise ratio and fragment ion abundance evaluation indicators are adopted to optimize the spectrum quality. Through spectrum smoothing technology, a final highly specific fragment ion spectrum is generated.

[0008] Optionally, according to the fragment ion spectrum, a multivariate data analysis algorithm based on machine learning is adopted, combined with the standard curve of the target compound and the interferent database, for quantitative analysis. Through dynamic interference correction technology, combined with principal component analysis and partial least squares method, a quantitative result is generated, including: According to the highly specific fragment ion spectrum, data preprocessing technology is adopted, combined with a feature extraction algorithm, to generate a preliminary quantitative feature data set; For the preliminary quantitative feature data set, dynamic interference correction technology is adopted, combined with the interferent database, to identify and correct interferent signals. Through principal component analysis and partial least squares method, a preliminary corrected data set is generated; For the preliminary corrected data set, a multivariate data analysis algorithm based on machine learning is adopted, combined with the standard curve of the target compound, for quantitative analysis. Through model optimization technology, a preliminary quantitative result is generated; For the preliminary quantitative result, accuracy evaluation indicators are adopted to optimize the quantitative result. Through an error correction mechanism, a final high-precision quantitative result is generated.

[0009] Another embodiment of the present application provides a liquid chromatography tandem mass spectrometry detection system, and the system includes: A design module, configured to design a multi-dimensional gradient elution program according to the physicochemical properties of the target compound, use the elution program to perform liquid chromatography separation on the sample, and combine a dynamic gradient optimization algorithm to adjust the mobile phase ratio and flow rate in real time to obtain a chromatographic separation spectrum; Ionization module, which is used to efficiently ionize target compounds according to the chromatographic separation spectrum, adopt ionization efficiency enhancement technology based on electrospray ionization, combine dynamic voltage tuning and solvent-assisted ionization strategies, dynamically adjust the ion source parameters through an adaptive electric field optimization algorithm, and obtain ionization signals; Screening module, which is used to fragment target ions according to the ionization signals, adopt multi-stage mass spectrometry fragmentation technology, combine a dynamic energy tuning algorithm, dynamically select characteristic fragment ions through fragment ion screening and combine the fragmentation rule library of target compounds to obtain fragment ion spectra; Analysis module, which is used to perform quantitative analysis according to the fragment ion spectra, adopt a multivariate data analysis algorithm based on machine learning, combine the standard curve of target compounds and the interferent database, and generate quantitative results through dynamic interference correction technology, combining principal component analysis and partial least squares method.

[0010] Another embodiment of the present application provides a storage medium, in which a computer program is stored. Wherein, the computer program is set to execute the method described in any one of the above when running.

[0011] Another embodiment of the present application provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and the processor is set to run the computer program to execute the method described in any one of the above.

[0012] Compared with the prior art, a liquid chromatography tandem mass spectrometry detection method provided by the present invention designs a multi-dimensional gradient elution program according to the physical and chemical properties of target compounds to obtain a chromatographic separation spectrum; efficiently ionizes target compounds according to the chromatographic separation spectrum to obtain ionization signals; fragments target ions according to the ionization signals, and obtains fragment ion spectra through fragment ion screening; performs quantitative analysis according to the fragment ion spectra, adopts a multivariate data analysis algorithm based on machine learning, combines the standard curve of target compounds and the interferent database, and generates quantitative results through dynamic interference correction technology, combining principal component analysis and partial least squares method. Therefore, it can improve the sensitivity, accuracy and real-time performance of detection through multi-dimensional gradient elution, dynamic ionization and fragmentation parameter optimization, and analysis means combining machine learning. Description of the Drawings

[0013] Figure 1 It is a hardware structure block diagram of a computer terminal for a liquid chromatography tandem mass spectrometry detection method provided by an embodiment of the present invention; Figure 2 It is a flow schematic diagram of a liquid chromatography tandem mass spectrometry detection method provided by an embodiment of the present invention; Figure 3Schematic structural diagram of a liquid chromatography tandem mass spectrometry detection system provided by an embodiment of the present invention. Detailed implementation manners

[0014] The embodiments described below by referring to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0015] An embodiment of the present invention first provides a liquid chromatography tandem mass spectrometry detection method, which can be applied to an electronic device, such as a computer terminal, specifically, a general computer, etc.

[0016] The following takes running on a computer terminal as an example to describe it in detail. Figure 1 Hardware structure block diagram of a computer terminal for a liquid chromatography tandem mass spectrometry detection method provided by an embodiment of the present invention. As Figure 1 shown, the computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the memory may include a non-volatile storage medium and an internal memory.

[0017] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any liquid chromatography tandem mass spectrometry detection method.

[0018] The processor is used to provide computing and control capabilities to support the operation of the entire computer device.

[0019] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any liquid chromatography tandem mass spectrometry detection method.

[0020] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art can understand that Figure 1 the structure shown in is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0021] It should be understood that the processor can be a Central Processing Unit (CPU), and the processor can also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0022] See Figure 2 , embodiments of the present invention provide a liquid chromatography tandem mass spectrometry detection method, which may include the following steps: S201, according to the physical and chemical properties of the target compound, design a multi-dimensional gradient elution program, use the elution program to perform liquid chromatography separation on the sample, and in combination with a dynamic gradient optimization algorithm, adjust the mobile phase ratio and flow rate in real time to obtain a chromatographic separation spectrum; In this step, by analyzing the physical and chemical properties such as the molecular structure, polarity, and solubility of the target compound, a multi-dimensional gradient elution program is formulated. This program optimizes the separation ability of the chromatographic column by adjusting the composition and flow rate of the mobile phase, so as to achieve efficient separation in liquid chromatography. The significance of designing a multi-dimensional gradient elution program is to improve the resolution and efficiency of chromatographic separation. By dynamically adjusting the mobile phase ratio and flow rate, better peak separation can be achieved in chromatographic analysis, reducing the co-elution phenomenon, and improving the sensitivity and accuracy of detection. This is particularly important for the analysis of complex samples because it can effectively separate the target compound, reduce interference, and improve the reliability of the analysis results.

[0023] Specifically, according to the physical and chemical properties of the target compound, a prediction model based on molecular descriptors can be used to calculate its retention time and elution conditions, and through a multi-objective optimization algorithm, a preliminary range of elution conditions can be generated; In this step, by analyzing the molecular descriptors of the target compound, its retention time in liquid chromatography is predicted, and in combination with a multi-objective optimization algorithm, a preliminary range of elution conditions is determined. The significance of calculating the retention time and elution conditions is to provide a reasonable starting point for chromatographic separation. Through the prediction model and optimization algorithm, the range of elution conditions can be effectively narrowed, improving the efficiency and accuracy of the experiment. This is particularly important for the analysis of complex samples because it can reduce the blindness of the experiment and improve the success rate of separation.

[0024] First, the system needs to collect the physicochemical property data of the target compound, which include molecular weight, polarity, solubility, boiling point, etc. Based on these data, the system can use molecular descriptors to characterize the properties of the compound. Molecular descriptors are the mathematical representations of compounds that can reflect their structures and properties. Commonly used molecular descriptors include topological indices, molecular fingerprints, and charge distributions, etc.

[0025] Next, the system uses a prediction model based on molecular descriptors to calculate the retention time and elution conditions of the target compound. Prediction models usually adopt machine learning algorithms such as support vector machine (SVM) or random forest (RF). These algorithms can handle high-dimensional data and provide good prediction performance. By training the model, the system can predict the retention time of the compound in liquid chromatography and determine the suitable elution conditions.

[0026] Finally, the system uses a multi-objective optimization algorithm to generate a preliminary range of elution conditions. The multi-objective optimization algorithm can consider multiple objectives simultaneously, such as resolution, analysis time, and solvent consumption, etc., to find the optimal elution conditions. Commonly used multi-objective optimization algorithms include genetic algorithm (GA) and particle swarm optimization (PSO). Through optimization, the system can determine a reasonable range of elution conditions, providing a basis for subsequent chromatographic separation.

[0027] Suppose the system is analyzing a novel drug molecule. First, the system collects the physicochemical property data of this molecule and uses molecular descriptors to characterize it. By training a random forest model, the system predicts that the retention time of this molecule in liquid chromatography is 10 minutes. Then, the system uses the genetic algorithm to optimize the elution conditions and finally determines a preliminary range of elution conditions: mobile phase A is water, mobile phase B is acetonitrile, the gradient elution time is 0 - 20 minutes, and the flow rate is 1 mL / min.

[0028] Another example, suppose the system is analyzing a novel drug with a molecular weight of 450 Da, medium polarity, and good water solubility. First, the system collects the physicochemical property data of this drug and uses molecular descriptors to characterize it. Molecular descriptors include molecular weight, polarity index, and solubility parameter, etc. Next, the system uses a support vector machine (SVM) model to predict the retention time of this drug in liquid chromatography. By training the model, the system predicts that the retention time of this drug is 12 minutes. To determine the suitable elution conditions, the system uses the genetic algorithm (GA) for multi-objective optimization. The optimization objectives include maximizing the resolution, minimizing the analysis time, and reducing the solvent consumption. Through the genetic algorithm, the system generates a preliminary range of elution conditions: mobile phase A is water, mobile phase B is methanol, the gradient elution time is 0 - 25 minutes, and the flow rate is 0.8 mL / min. This range provides a reasonable starting point for subsequent chromatographic separation.

[0029] For the initial elution condition range, a multi-dimensional gradient elution program generation algorithm is adopted. Combining the three-dimensional parameter space of time-mobile phase ratio-flow rate, a gradient elution program is designed. Through dynamic constraint conditions, an initial elution program is generated. This step optimizes the chromatographic separation conditions by designing a gradient elution program in the three-dimensional parameter space to achieve efficient separation in liquid chromatography. The role of designing the gradient elution program is to improve the resolution and efficiency of chromatographic separation. By optimizing the elution program in the three-dimensional parameter space, better peak separation can be achieved, co-elution phenomenon can be reduced, and the sensitivity and accuracy of detection can be improved. This is particularly important for the analysis of complex samples because it can effectively separate the target compounds, reduce interference, and improve the reliability of the analysis results.

[0030] After determining the initial elution condition range, the system needs to design a detailed gradient elution program. The design of the gradient elution program involves three dimensions: time, mobile phase ratio, and flow rate. The system first defines an initial point in this three-dimensional parameter space, and this point represents the initial elution conditions.

[0031] Then, the system uses a multi-dimensional gradient elution program generation algorithm to explore the parameter space. This algorithm generates a series of elution programs by adjusting time, mobile phase ratio, and flow rate. The system can use the simulated annealing algorithm (SA) or Latin hypercube sampling (LHS) to explore the parameter space to find the optimal elution program.

[0032] During the process of generating the elution program, the system needs to consider dynamic constraint conditions. These constraint conditions include the pressure resistance of the chromatographic column, the compatibility of the mobile phase, and the limitation of the analysis time, etc. By dynamically adjusting these constraint conditions, the system can ensure that the generated elution program is feasible in actual operation.

[0033] When analyzing a novel drug molecule, the system defines an initial point in the three-dimensional parameter space of time-mobile phase ratio-flow rate: the time is 0 minutes, the mobile phase ratio is 50% water and 50% acetonitrile, and the flow rate is 1 mL / min. Through the simulated annealing algorithm, the system generates a series of elution programs and selects an optimal program: within 0 - 10 minutes, the mobile phase ratio linearly increases from 50% water and 50% acetonitrile to 20% water and 80% acetonitrile, and the flow rate remains at 1 mL / min.

[0034] After determining the initial elution condition range, the system needs to design a detailed gradient elution program. First, the system defines an initial point in the three-dimensional parameter space of time-mobile phase ratio-flow rate: the time is 0 minutes, the mobile phase ratio is 60% water and 40% methanol, and the flow rate is 0.8 mL / min. The system uses the simulated annealing algorithm (SA) to explore the parameter space. By adjusting the time, mobile phase ratio, and flow rate, the system generates a series of elution programs. To ensure the feasibility of the program, the system sets dynamic constraints, such as the maximum pressure tolerance of the chromatographic column is 400 bar, and the maximum ratio of the mobile phase is 90% methanol. After multiple iterations, the system selects an optimal elution program: within 0-10 minutes, the mobile phase ratio linearly increases from 60% water and 40% methanol to 20% water and 80% methanol, and the flow rate remains at 0.8 mL / min. This program was verified in the laboratory and showed good separation effects.

[0035] According to the preliminary elution program, the dynamic gradient optimization algorithm is adopted, combined with the real-time chromatographic peak shape monitoring data, to adjust the mobile phase ratio and flow rate. Through the feedback control mechanism, the separation effect is optimized to generate a preliminary high-resolution chromatographic separation spectrum. This step optimizes the chromatographic separation effect by monitoring the chromatographic peak shape in real time and dynamically adjusting the elution conditions. The significance of optimizing the separation effect is to improve the resolution and efficiency of chromatographic separation. By dynamically adjusting the elution conditions, better peak separation can be achieved in chromatographic analysis, reducing the co-elution phenomenon, and improving the sensitivity and accuracy of detection. This is particularly important for the analysis of complex samples because it can effectively separate the target compounds, reduce interference, and improve the reliability of the analysis results.

[0036] After designing the preliminary elution program, the system needs to optimize the separation effect in actual operation. For this purpose, the system adopts the dynamic gradient optimization algorithm, which can dynamically adjust the mobile phase ratio and flow rate according to the real-time chromatographic peak shape monitoring data. The system installs real-time monitoring equipment in the liquid chromatography system, which can collect chromatographic peak shape data in real time, such as peak height, peak width, and symmetry. By analyzing these data, the system can judge the current separation effect and adjust the elution conditions accordingly.

[0037] The feedback control mechanism plays a key role in the optimization process. The system can use a proportional-integral-derivative (PID) controller to achieve feedback control. The PID controller optimizes the chromatographic separation effect by adjusting the mobile phase ratio and flow rate. Through multiple iterations, the system can generate a preliminary high-resolution chromatographic separation spectrum.

[0038] In actual operation, the system uses a dynamic gradient optimization algorithm to optimize the elution program of new drug molecules. Through a real-time monitoring device, the system discovers that the symmetry of some chromatographic peaks is poor. Using a PID controller, the system adjusts the mobile phase ratio and flow rate, and finally obtains a high-resolution chromatographic separation spectrum in which the symmetry and resolution of all chromatographic peaks are significantly improved.

[0039] In the laboratory, the system uses a liquid chromatography system to separate new drugs. To optimize the separation effect, the system installs a real-time monitoring device that can collect chromatographic peak shape data in real time, such as peak height, peak width, and symmetry. The system adopts a dynamic gradient optimization algorithm and adjusts the mobile phase ratio and flow rate according to the real-time monitoring data. By analyzing the chromatographic peak shape, the system discovers that the symmetry of some peaks is poor, which may affect the separation effect. Using a proportional-integral-derivative (PID) controller, the system dynamically adjusts the mobile phase ratio and flow rate. After multiple adjustments, the system obtains a preliminary high-resolution chromatographic separation spectrum. This spectrum shows good peak separation and symmetry, and all chromatographic peaks of the target compounds are clearly separated.

[0040] For the preliminary high-resolution chromatographic separation spectrum, a peak shape optimization algorithm is adopted, combined with baseline correction and noise filtering techniques, to further improve the spectrum quality. Through resolution evaluation indicators, a final high-resolution chromatographic separation spectrum is generated.

[0041] This step improves the quality of the chromatographic separation spectrum by optimizing the chromatographic peak shape and reducing noise. The role of improving the spectrum quality is to improve the accuracy and reliability of chromatographic analysis. By optimizing the chromatographic peak shape and reducing noise, a clearer chromatographic separation spectrum can be obtained, improving the detection sensitivity and accuracy. This is particularly important for the analysis of complex samples because it can provide more reliable analysis results to support subsequent qualitative and quantitative analysis.

[0042] After obtaining the preliminary high-resolution chromatographic separation spectrum, the system needs to further improve the spectrum quality. For this purpose, the system adopts a peak shape optimization algorithm that can optimize the shape of chromatographic peaks and improve the resolution. The system first performs baseline correction to eliminate the baseline shift caused by instrument drift or background noise. Common baseline correction methods include polynomial fitting and moving average method. Next, the system uses noise filtering techniques to remove random noise in the spectrum, improving the clarity and accuracy of the signal. Common noise filtering techniques include wavelet transform and Fourier transform. Finally, the system uses resolution evaluation indicators to evaluate the quality of the spectrum. Resolution evaluation indicators include peak resolution, peak capacity, and signal-to-noise ratio, etc. By optimizing these indicators, the system can generate a final high-resolution chromatographic separation spectrum.

[0043] When improving the chromatographic separation profile quality of novel drug molecules, the system first performs baseline correction using polynomial fitting method, eliminating baseline drift. Then, the system uses wavelet transform to remove random noise in the profile. By optimizing the peak shape, the system improves the resolution and symmetry of chromatographic peaks. Finally, the system uses a resolution evaluation metric to evaluate the quality of the profile and confirms that it meets the high-resolution standard.

[0044] After obtaining a preliminary high-resolution chromatographic separation profile, the system needs to further improve the profile quality. First, the system performs baseline correction using polynomial fitting method, eliminating baseline offsets caused by instrument drift or background noise. Then, the system uses wavelet transform to remove random noise in the profile, improving the clarity and accuracy of the signal. To optimize the chromatographic peak shape, the system adopts a peak shape optimization algorithm to adjust the peak width and symmetry. Finally, the system uses a resolution evaluation metric to evaluate the quality of the profile. By optimizing the peak resolution, peak capacity, and signal-to-noise ratio, the system generates the final high-resolution chromatographic separation profile. This profile shows excellent separation performance, providing a reliable data basis for subsequent qualitative and quantitative analysis.

[0045] S202, according to the chromatographic separation profile, adopt an ionization efficiency enhancement technique based on electrospray ionization, combine dynamic voltage tuning and solvent-assisted ionization strategies to efficiently ionize the target compound, and dynamically adjust the ion source parameters through an adaptive electric field optimization algorithm to obtain an ionization signal; This step improves the ionization efficiency of the target compound by optimizing the electrospray ionization conditions, ensuring a strong and stable ionization signal in mass spectrometry analysis. The role of efficient ionization is to improve the sensitivity and signal intensity of mass spectrometry detection. By optimizing the ionization conditions, higher ionization efficiency can be obtained, thereby improving the detection limit and quantitative accuracy of the target compound. This is particularly important for the analysis of low-concentration samples as it enables accurate detection and quantification of the target compound in a complex matrix.

[0046] Specifically, according to the high-resolution chromatographic separation profile and combined with the physicochemical properties of the target compound, initialize the electrospray ionization conditions and generate preliminary ionization parameters through an ionization efficiency prediction model; This step determines the initial conditions of electrospray ionization by analyzing the physicochemical properties of the target compound to achieve efficient ionization in mass spectrometry analysis. The significance of initializing the electrospray ionization conditions is to provide a reasonable starting point for mass spectrometry analysis. Through the prediction model and optimization algorithm, the ionization efficiency can be effectively improved, ensuring a strong and stable ionization signal in mass spectrometry analysis. This is particularly important for the analysis of low-concentration samples as it can improve the detection sensitivity and signal intensity.

[0047] After liquid chromatography separation, the system obtained a high-resolution chromatogram. Next, the system needs to initialize the conditions for the electrospray ionization process of the target compound. First, the system analyzes the physical and chemical properties of the target compound, including molecular weight, polarity, acidity / basicity, and solubility, etc. These properties will affect the behavior of the compound during the electrospray ionization process. To improve the ionization efficiency, the system uses an ionization efficiency prediction model. This model is based on a large amount of experimental data and can predict the ionization efficiency of different compounds under specific conditions. By inputting the physical and chemical properties of the target compound, the model outputs preliminary ionization parameters such as spray voltage, solvent ratio, and flow rate, etc. For example, assume the target compound is a small molecule drug with medium polarity. Through model prediction, the system recommends an initial spray voltage of 3.5 kV, a solvent ratio of 70% methanol and 30% water, and a flow rate of 0.3 mL / min. These parameters provide a reasonable starting point for the subsequent ionization process.

[0048] Again, after liquid chromatography separation, the system obtained a high-resolution chromatogram showing a clear peak shape of the target compound. Assume the target compound is a new antibiotic with high polarity and medium molecular weight. The system first analyzed the physical and chemical properties of this compound, including its molecular weight, polarity, acidity / basicity, and solubility, etc. To improve the ionization efficiency, the system used a machine learning-based ionization efficiency prediction model. This model has been trained with a large amount of experimental data and can predict the efficiency of the compound during the electrospray ionization process based on its physical and chemical properties. The system input the properties of the target compound, and the model output preliminary ionization parameter recommendations: a spray voltage of 3.5 kV, a solvent ratio of 80% methanol and 20% water, and a flow rate of 0.2 mL / min. These parameters provide a reasonable starting point for the subsequent ionization process.

[0049] When analyzing a new antibiotic (for example, compound Y), the system first obtained a high-resolution chromatogram. Compound Y has high polarity and a specific molecular weight (such as 350 Da). The system combined the physical and chemical properties of compound Y to initialize the electrospray ionization (ESI) conditions, including parameters such as spray voltage, capillary temperature, and gas flow rate. To optimize these parameters, the system used a neural network-based ionization efficiency prediction model. This model can predict the optimal ionization conditions based on the physical and chemical properties of the compound (such as polarity, molecular weight). Through model prediction, the system generated preliminary ionization parameters to ensure an efficient ionization signal can be obtained during the subsequent ionization process.

[0050] For the preliminary ionization parameters, a dynamic voltage tuning algorithm is adopted, combined with a solvent-assisted ionization strategy, to adjust the spray voltage and solvent ratio in real time. Through ionization efficiency monitoring technology, preliminary optimized ionization conditions are generated; This step improves the ionization efficiency of the target compound by dynamically adjusting the ionization conditions. The role of optimizing the ionization conditions is to enhance the sensitivity and signal intensity of mass spectrometry detection. By optimizing the ionization conditions, higher ionization efficiency can be achieved, thereby improving the detection limit and quantification accuracy of the target compound. This is particularly important for the analysis of low-concentration samples as it ensures the accurate detection and quantification of the target compound in complex matrices.

[0051] After obtaining the preliminary ionization parameters, the system needs to further optimize these conditions to improve the ionization efficiency. To this end, the system employs a dynamic voltage tuning algorithm, which can dynamically adjust the spray voltage and solvent ratio based on the ionization efficiency data monitored in real time. The system installs an ionization efficiency monitoring device in the mass spectrometer, which can measure the intensity and stability of the ion signal in real time. By analyzing this data, the system can judge the current ionization efficiency and accordingly adjust the spray voltage and solvent ratio. Combining with the solvent-assisted ionization strategy, the system can affect the ionization efficiency by changing the composition of the solvent. For example, increasing the proportion of organic solvent can improve the ionization efficiency of non-polar compounds. After multiple adjustments, the system generates preliminary optimized ionization conditions to ensure strong and stable ion signals in mass spectrometry analysis.

[0052] During the experiment, the system applies a dynamic voltage tuning algorithm (e.g., genetic algorithm) to optimize the ionization process. The system combines the solvent-assisted ionization strategy and selects a suitable solvent (such as a mixture of methanol and water) to improve the ionization efficiency. By monitoring the ionization efficiency in real time, the system adjusts the spray voltage and solvent ratio to obtain the best ionization effect. Ionization efficiency monitoring techniques (such as on-line mass spectrometry monitoring) help the system to evaluate the effectiveness of the ionization conditions in real time. Finally, the system generates preliminary optimized ionization conditions to ensure the intensity and stability of the ionization signal.

[0053] For the preliminary optimized ionization conditions, an adaptive electric field optimization algorithm is adopted, combined with ion signal intensity feedback, to dynamically adjust the ion source parameters. Through the electric field distribution simulation technology, preliminary high-quality ionization signals are generated; This step improves the quality of the ionization signal by optimizing the electric field distribution. The significance of generating high-quality ionization signals lies in improving the accuracy and reliability of mass spectrometry analysis. By optimizing the electric field distribution, stronger ion signals can be obtained, improving the detection sensitivity and signal intensity. This is particularly important for the analysis of low-concentration samples as it ensures the accurate detection and quantification of the target compound in complex matrices.

[0054] After optimizing the ionization conditions, the system needs to ensure that the generated ion signals have high quality. To this end, the system adopts an adaptive electric field optimization algorithm, which can dynamically adjust the ion source parameters according to the feedback of the ion signal intensity. The system uses electric field distribution simulation technology to optimize the electric field distribution of the ion source. Through simulation, the system can identify the non-uniform regions in the electric field and adjust the geometric structure or voltage settings of the ion source to improve the electric field distribution. This helps to improve the ionization efficiency and signal intensity. During the experiment, the system monitors the intensity and stability of the ion signals in real time. Through multiple iterations and adjustments, the system generates preliminary high-quality ionization signals. These signals have a high signal-to-noise ratio and good stability, providing a reliable data basis for subsequent mass spectrometry analysis.

[0055] The system uses an adaptive electric field optimization algorithm (such as the adaptive genetic algorithm) to further optimize the ionization conditions. The system combines the ion signal intensity feedback to dynamically adjust the ion source parameters, such as the spray voltage and capillary position. Through the electric field distribution simulation technology (such as finite element analysis), the system can simulate and optimize the electric field distribution to improve the ionization efficiency and signal quality. The preliminary high-quality ionization signals show a high signal-to-noise ratio and stable ion intensity, providing a reliable data basis for subsequent mass spectrometry analysis.

[0056] For the preliminary high-quality ionization signals, signal-to-noise ratio and ion intensity stability evaluation indicators are adopted to optimize the signal quality, and the final high-quality ionization signals are generated through signal filtering technology.

[0057] This step improves the accuracy and reliability of mass spectrometry analysis by optimizing the signal quality. The role of optimizing the signal quality is to improve the accuracy and reliability of mass spectrometry analysis. By optimizing the signal quality, stronger ion signals can be obtained, improving the detection sensitivity and signal intensity. This is particularly important for the analysis of low-concentration samples, as it can ensure the accurate detection and quantification of target compounds in complex matrices.

[0058] After obtaining the preliminary high-quality ionization signals, the system needs to further optimize the signal quality. First, the system uses the signal-to-noise ratio and ion intensity stability as evaluation indicators. The signal-to-noise ratio is an important indicator to measure the signal quality, and a higher signal-to-noise ratio means a higher proportion of useful information in the signal. The ion intensity stability reflects the stability and repeatability of the signal. The system adopts signal filtering technology to remove the random noise in the signal. Commonly used filtering technologies include wavelet transform and Fourier transform, which can effectively remove the noise and improve the clarity and accuracy of the signal. Through multiple optimizations and adjustments, the system generates the final high-quality ionization signals. These signals have a high signal-to-noise ratio and good stability, providing a reliable basis for subsequent multi-stage mass spectrometry analysis. This process ensures the accuracy and sensitivity of mass spectrometry analysis, especially in the analysis of complex samples, and can provide more reliable results.

[0059] To further improve the signal quality, the system optimizes the preliminary high-quality ionization signals. The system uses the signal-to-noise ratio (SNR) and ion intensity stability as evaluation metrics to identify noise and unstable factors in the signals. By applying signal filtering techniques (such as Kalman filtering or wavelet transform), the system can reduce noise and improve the clarity and stability of the signals. The finally generated high-quality ionization signals have a higher signal-to-noise ratio and more stable ion intensity, providing a reliable data basis for quantitative analysis and compound identification.

[0060] S203, based on the ionization signal, adopts the multi-stage mass spectrometry fragmentation technique, combines with the dynamic energy tuning algorithm to fragment the target ions. Through fragment ion screening and combining with the fragmentation rule library of the target compound, characteristic fragment ions are dynamically selected to obtain the fragment ion spectrum; This step selectively fragments the target ions through multi-stage mass spectrometry technology to generate a characteristic fragment ion spectrum for subsequent qualitative and quantitative analysis. The significance of multi-stage mass spectrometry fragmentation lies in improving the specificity and accuracy of compound identification. By generating a characteristic fragment ion spectrum, the target compound can be more accurately identified, reducing the possibility of misidentification. This is particularly important for the analysis of complex samples as it can provide more detailed structural information to support the qualitative and quantitative analysis of compounds.

[0061] Specifically, according to the high-quality ionization signal, combined with the molecular weight and structural characteristics of the target compound, the fragmentation energy is initialized, and through the fragmentation efficiency prediction model, preliminary fragmentation parameters are generated; This step determines the initial conditions of the fragmentation energy by analyzing the molecular weight and structural characteristics of the target compound to achieve efficient fragmentation in mass spectrometry analysis. The significance of initializing the fragmentation energy is to provide a reasonable starting point for mass spectrometry analysis. Through the prediction model and optimization algorithm, the fragmentation efficiency can be effectively improved to ensure strong and stable fragment ion signals in mass spectrometry analysis. This is particularly important for the analysis of complex samples as it can improve the detection sensitivity and signal intensity.

[0062] After obtaining a high-quality ionization signal, the system needs to initialize the energy parameters for the mass spectrometry fragmentation process of the target compound. Assume that the target compound is a complex organic molecule with multiple functional groups and a large molecular weight. The system first analyzes the molecular weight and structural characteristics of the compound to determine its possible fragmentation paths and energy requirements. To improve the fragmentation efficiency, the system uses a fragmentation efficiency prediction model. This model is based on a large amount of experimental data and can predict the fragmentation efficiency of different compounds under specific energy conditions. The system inputs the molecular weight and structural characteristics of the target compound, and the model outputs preliminary fragmentation parameter suggestions: the collision energy is 35 eV. This parameter provides a reasonable starting point for the subsequent fragmentation process.

[0063] When analyzing a new drug (e.g., Compound X), the system first obtains a high-quality ionization signal of the drug. The molecular weight of Compound X is 500 Da, and it has a complex cyclic structure and multiple functional groups. The system uses the molecular weight and structural characteristics of Compound X to initialize the fragmentation energy and selects a suitable collision energy range (e.g., 20 - 40 eV). To optimize the fragmentation parameters, the system uses a fragmentation efficiency prediction model based on support vector machine (SVM). This model can predict the optimal fragmentation energy according to the structural characteristics of the compound (such as molecular weight, functional group type). Through model prediction, the system generates preliminary fragmentation parameters to ensure an efficient fragmentation effect in subsequent mass spectrometry analysis.

[0064] For the preliminary fragmentation parameters, a dynamic energy tuning algorithm is adopted, combined with the fragment ion intensity feedback, to adjust the collision energy in real time. Through a multi-stage fragmentation strategy, a preliminary fragment ion spectrum is generated; This step improves the fragmentation efficiency of the target compound by dynamically adjusting the fragmentation energy. The role of generating the fragment ion spectrum is to improve the specificity and accuracy of compound identification. By generating a characteristic fragment ion spectrum, the target compound can be more accurately identified, reducing the possibility of misidentification. This is particularly important for the analysis of complex samples because it can provide more detailed structural information to support the qualitative and quantitative analysis of the compound.

[0065] After obtaining the preliminary fragmentation parameters, the system needs to further optimize these parameters to improve the generation efficiency of fragment ions. To this end, the system adopts a dynamic energy tuning algorithm, which can dynamically adjust the collision energy according to the fragment ion intensity data monitored in real time. The system installs a fragment ion intensity monitoring device in the mass spectrometer, which can measure the intensity and stability of fragment ions in real time. By analyzing these data, the system can judge the current fragmentation efficiency and adjust the collision energy accordingly. Combining with the multi-stage fragmentation strategy, the system can generate different fragment ions by gradually increasing the collision energy. This strategy helps to comprehensively characterize the fragmentation path of the target compound. After multiple adjustments, the system generates a preliminary fragment ion spectrum, showing the main fragment ions of the target compound.

[0066] During the experiment, the system applies a dynamic energy tuning algorithm (e.g., particle swarm optimization algorithm) to optimize the fragmentation process. The system monitors the intensity feedback of fragment ions in real time and adjusts the collision energy to obtain the best fragmentation effect. Through the multi-stage fragmentation strategy, the system can fragment the target ions multiple times at different energy levels, thus generating richer fragment ion information. The preliminary fragment ion spectrum shows multiple characteristic fragment ion peaks, which can provide detailed information about the structure of compound X and lay a foundation for subsequent analysis.

[0067] For the preliminary fragment ion spectrum, a screening algorithm based on the fragmentation rule library is adopted. Combining with the characteristic fragments of the target compound, characteristic fragment ions are dynamically selected. Through the fragment ion matching technology, a preliminary highly specific fragment ion spectrum is generated; This step improves the specificity of compound identification by selecting characteristic fragment ions. The significance of selecting characteristic fragment ions lies in improving the specificity and accuracy of compound identification. By selecting characteristic fragment ions, the target compound can be more accurately identified, reducing the possibility of misidentification. This is particularly important for the analysis of complex samples, as it can provide more detailed structural information to support the qualitative and quantitative analysis of compounds.

[0068] After obtaining the preliminary fragment ion spectrum, the system needs to select characteristic fragment ions from it for more precise analysis. First, the system uses a screening algorithm based on the fragmentation rule library. This algorithm contains the known fragmentation paths and characteristic fragment information of a large number of compounds. The system compares the preliminary fragment ion spectrum with the fragmentation rule library to identify the ions that match the characteristic fragments of the target compound. Through this comparison, the system can dynamically select the most specific fragment ions. To ensure that the selected fragment ions have high specificity, the system uses fragment ion matching technology. This technology can accurately match the mass and abundance of fragment ions to ensure that the selected ions can represent the characteristics of the target compound. Finally, the system generates a preliminary highly specific fragment ion spectrum.

[0069] The system uses a database (e.g., MassBank) containing the fragmentation rules of known compounds to screen the preliminary fragment ion spectra. The system applies screening algorithms (such as decision tree algorithms) to identify the characteristic fragment ions related to compound X. These characteristic fragments are usually the markers of the key structural units in the molecule of compound X. Through fragment ion matching techniques (such as matching based on mass-to-charge ratio), the system can confirm the identities of these characteristic fragments and generate a preliminary highly specific fragment ion spectrum. This spectrum has high specificity and can effectively distinguish compound X from other possible interferents.

[0070] For the preliminary highly specific fragment ion spectrum, signal-to-noise ratio and fragment ion abundance evaluation indicators are used to optimize the spectrum quality, and through spectrum smoothing techniques, a final highly specific fragment ion spectrum is generated.

[0071] This step improves the accuracy and reliability of compound identification by optimizing the spectrum quality. The role of optimizing the spectrum quality is to improve the accuracy and reliability of compound identification. By optimizing the spectrum quality, a clearer fragment ion spectrum can be obtained, improving the detection sensitivity and accuracy. This is particularly important for the analysis of complex samples as it can provide more reliable analysis results to support subsequent qualitative and quantitative analysis.

[0072] After obtaining the preliminary highly specific fragment ion spectrum, the system needs to further optimize the spectrum quality. First, the system uses signal-to-noise ratio and fragment ion abundance as evaluation indicators. Signal-to-noise ratio is an important indicator for measuring the spectrum quality, and a higher signal-to-noise ratio means a higher proportion of useful information in the spectrum. Fragment ion abundance reflects the intensity and stability of the ions in the spectrum. The system adopts spectrum smoothing techniques to remove the random noise in the spectrum. Commonly used smoothing techniques include Gaussian smoothing and moving average, and these techniques can effectively remove the noise and improve the clarity and accuracy of the spectrum. Through multiple optimizations and adjustments, the system generates the final highly specific fragment ion spectrum. This spectrum has a high signal-to-noise ratio and good ion abundance, providing a reliable basis for subsequent quantitative analysis. This process ensures the accuracy and sensitivity of mass spectrometry analysis, especially in the analysis of complex samples, and can provide more reliable results.

[0073] To further improve the spectrum quality, the system optimizes the preliminary highly specific fragment ion spectrum. The system uses signal-to-noise ratio (SNR) and fragment ion abundance as evaluation indicators to identify the noise and low-abundance fragments in the spectrum. By applying spectrum smoothing techniques (such as Savitzky-Golay smoothing or Gaussian smoothing), the system can reduce the noise and improve the clarity of the signal. The finally generated highly specific fragment ion spectrum has a higher signal-to-noise ratio and clearer characteristic fragment peaks, providing a reliable data basis for quantitative analysis and compound identification.

[0074] S204. Based on the fragment ion spectrum, a multivariate data analysis algorithm based on machine learning is adopted, combined with the standard curve of the target compound and the interference database, for quantitative analysis. Through the dynamic interference correction technology, combined with principal component analysis and partial least squares method, the quantitative results are generated.

[0075] In this step, the fragment ion spectrum is analyzed by a machine learning algorithm, combined with the standard curve and the interference database, for accurate quantitative analysis. The role of quantitative analysis is to improve the accuracy and reliability of the analysis results. By combining the machine learning algorithm and the dynamic interference correction technology, interference signals can be effectively identified and corrected, improving the accuracy of the quantitative results. This is particularly important for the analysis of complex samples, as it can ensure the accurate quantification of the target compound in the presence of interferents.

[0076] Specifically, based on the highly specific fragment ion spectrum, data preprocessing techniques can be adopted, combined with feature extraction algorithms, to generate a preliminary quantitative feature dataset. In this step, the fragment ion spectrum is preprocessed and features are extracted to generate a dataset for quantitative analysis. The significance of generating the quantitative feature dataset is to provide a reliable data basis for quantitative analysis. Through data preprocessing and feature extraction, the quality and accuracy of the data can be effectively improved, ensuring reliable results in quantitative analysis. This is particularly important for the analysis of complex samples, as it can provide more accurate quantitative results to support subsequent analysis and research.

[0077] In mass spectrometry analysis, data preprocessing is a key step to ensure the accuracy of the analysis. First, the system performs noise removal and baseline correction on the highly specific fragment ion spectrum. Noise removal usually uses wavelet transform or Fourier transform, which can effectively remove high-frequency noise and retain useful signals. Baseline correction is achieved through polynomial fitting or locally weighted regression (LOWESS) to eliminate the influence of baseline drift on the signal. Through these preprocessing steps, the signal-to-noise ratio of the spectrum is significantly improved, laying a foundation for subsequent feature extraction.

[0078] Next, the system uses feature extraction algorithms to extract characteristic ions related to the target compound from the preprocessed spectrum. Commonly used feature extraction methods include principal component analysis (PCA) and independent component analysis (ICA). PCA identifies the most representative features in the data through dimensionality reduction techniques, while ICA is used to separate independent signal components. These characteristic ions are usually the characteristic fragments of the target compound and can provide information about the compound concentration. Through these steps, the system generates a preliminary quantitative feature dataset that contains the key feature information of the target compound.

[0079] Finally, the system standardizes the extracted features to ensure comparability between different features. Standardization usually includes mean normalization or min-max normalization, which can adjust the feature values to the same scale range and eliminate the bias caused by different dimensions. The standardized feature dataset provides a reliable data basis for subsequent quantitative analysis.

[0080] When analyzing the concentration of a new antibiotic, the system first obtains the high-specificity fragment ion spectrum of the antibiotic. Since mass spectrometry data usually contains a large amount of noise and baseline drift, the system uses wavelet transform to remove high-frequency noise. This process is achieved by decomposing the signal and removing irrelevant high-frequency components, thereby improving the signal-to-noise ratio. Next, the system applies polynomial fitting for baseline correction to eliminate the influence of baseline drift on the signal. After the preprocessing is completed, the system uses principal component analysis (PCA) to extract the characteristic ions related to the antibiotic from the spectrum. PCA reduces the dimensionality by identifying the most representative features in the data and extracts the key ions that can reflect the antibiotic concentration. These characteristic ions appear as significant peaks in the spectrum and can provide information about the antibiotic concentration. Finally, the system generates a preliminary quantitative feature dataset, which contains the key feature information of the antibiotic and lays the foundation for subsequent quantitative analysis.

[0081] For the preliminary quantitative feature dataset, the dynamic interference correction technique is adopted, combined with the interference substance database, to identify and correct the interference substance signal, and through principal component analysis and partial least squares method, a preliminary corrected dataset is generated; This step improves the accuracy of quantitative analysis by identifying and correcting the interference substance signal. The role of correcting the interference substance signal is to improve the accuracy and reliability of quantitative analysis. By identifying and correcting the interference substance signal, the influence of interference on the quantitative result can be effectively reduced and the analysis precision can be improved. This is particularly important for the analysis of complex samples because it can ensure the accurate quantification of the target compound in the presence of interference substances.

[0082] In quantitative analysis, the interference substance signal may affect the quantitative result of the target compound. To solve this problem, the system uses the dynamic interference correction technique. First, the system compares the preliminary quantitative feature dataset with the interference substance database to identify the possible interference substance signals. The interference substance database usually contains the characteristic ion information of known interference substances and can help identify the non-target signals in the spectrum.

[0083] Next, the system applies principal component analysis (PCA) and partial least squares (PLS) to separate the target compound signal and the interferent signal. PCA is used for dimensionality reduction and removing redundant information, while PLS is used to establish a linear relationship model between the target compound and the interferent. Through these steps, the system can effectively separate the signal of the target compound and reduce the influence of the interferent on the quantitative results.

[0084] Finally, the system verifies the corrected data set to ensure the correction effect of the interferent signal. The verification process usually includes comparing the signal intensities and characteristic ion abundances before and after correction to evaluate the accuracy and effectiveness of the correction. The verified corrected data set provides more accurate data for subsequent quantitative analysis.

[0085] In the analysis of antibiotic concentration, the system found that there might be solvent residues and other compounds in the samples, and these interferents might affect the quantitative results of antibiotics. To identify and correct these interferent signals, the system first compared the preliminary quantitative feature data set with a database containing the characteristic ion information of known interferents. Through the comparison, the system identified the characteristic ions related to solvent residues in the spectrum. Next, the system applied principal component analysis (PCA) to reduce the dimension and remove redundant information, identifying the main variation patterns in the data. Then, the system used partial least squares (PLS) to establish a linear relationship model between the target compound (antibiotic) and the interferent. PLS can effectively separate the target compound signal and the interferent signal. Through modeling and correction, the system successfully reduced the influence of the interferent on the quantitative results. After these steps, the system generated a corrected data set. To verify the correction effect, the system compared the signal intensities and characteristic ion abundances before and after correction, confirming that the interferent signal was effectively corrected. The corrected data set provided a more accurate data basis for subsequent quantitative analysis.

[0086] For the preliminary corrected data set, a machine learning-based multivariate data analysis algorithm is adopted, combined with the standard curve of the target compound, to conduct quantitative analysis. Through model optimization techniques, preliminary quantitative results are generated; This step analyzes the corrected data set through machine learning algorithms to generate quantitative results. The significance of conducting quantitative analysis is to improve the accuracy and reliability of the analysis results. By combining machine learning algorithms and standard curves, the precision and accuracy of the quantitative results can be effectively improved. This is particularly important for the analysis of complex samples because it can provide more reliable quantitative results to support subsequent analysis and research.

[0087] Based on the calibration dataset, the system uses machine learning algorithms for quantitative analysis. Commonly used algorithms include Support Vector Machine (SVM), Random Forest (RF), and Neural Network (NN). These algorithms can handle complex multivariate data and provide high-precision quantitative results. The system first compares the calibration dataset with the standard curve of the target compound, and the standard curve provides the response signal at known concentrations.

[0088] Next, the system trains a machine learning model to predict the concentration of unknown samples. The training process of the model includes selecting appropriate features, adjusting model parameters, and validating the accuracy of the model. Through cross-validation and model optimization techniques, the system ensures the accuracy and robustness of the model. Cross-validation can effectively evaluate the generalization ability of the model, while model optimization improves the prediction performance of the model by adjusting hyperparameters.

[0089] Finally, the system evaluates the prediction results of the model to ensure the accuracy of quantitative analysis. The evaluation process usually includes calculating the prediction error and the coefficient of determination, and analyzing the prediction performance and stability of the model. The evaluated model can provide reliable quantitative results, providing a basis for subsequent precision optimization.

[0090] Based on the calibration dataset, the system selects the Random Forest (RF) algorithm for the quantitative analysis of antibiotics. First, the system compares the calibrated dataset with the standard curve of antibiotics. The standard curve is drawn from the response signals measured from antibiotic samples with known concentrations, providing the relationship between concentration and signal intensity. The system uses the calibration dataset to train the Random Forest model. The training process includes selecting appropriate features, adjusting model parameters, and validating the accuracy of the model. Through cross-validation, the system evaluates the generalization ability of the model and optimizes the prediction performance of the model by adjusting hyperparameters (such as the number and depth of trees). After training and optimization, the Random Forest model can accurately predict the antibiotic concentration in unknown samples. The system evaluates the prediction results of the model, calculates the prediction error and the coefficient of determination, and confirms the prediction performance and stability of the model. The preliminary quantitative results show the antibiotic concentration in the sample, providing a basis for subsequent precision optimization.

[0091] For the preliminary quantitative results, precision evaluation metrics are adopted to optimize the quantitative results, and through an error correction mechanism, final high-precision quantitative results are generated.

[0092] This step improves the accuracy and reliability of the analysis by optimizing the quantitative results. The role of optimizing the quantitative results is to improve the accuracy and reliability of the analysis results. By optimizing the quantitative results, the precision and accuracy of quantitative analysis can be effectively improved. This is particularly important for the analysis of complex samples because it can provide more reliable quantitative results to support subsequent analysis and research.

[0093] After obtaining the preliminary quantitative results, the system needs to optimize and evaluate the accuracy of the results. First, the system uses accuracy evaluation metrics such as mean squared error (MSE) and coefficient of determination (R²) to evaluate the prediction performance of the model. The mean squared error is used to measure the average deviation between the predicted value and the true value, while the coefficient of determination reflects the ability of the model to explain the data.

[0094] If a large error is found, the system will apply error correction mechanisms such as bias correction and residual analysis to further optimize the quantitative results. Bias correction improves the prediction accuracy by adjusting the systematic error of the model, while residual analysis is used to identify and correct outliers in the model. Through these steps, the system can identify and correct the systematic errors in the model, ensuring the accuracy and reliability of the quantitative results.

[0095] Finally, the system verifies the optimized quantitative results to ensure their accuracy and reliability. The verification process usually includes comparative analysis with known standard samples to evaluate the accuracy and consistency of the quantitative results. The verified high-precision quantitative results provide a solid foundation for scientific research and applications.

[0096] After obtaining the preliminary quantitative results, the system uses accuracy evaluation metrics such as mean squared error (MSE) and coefficient of determination (R²) to evaluate the prediction performance of the model. The mean squared error is used to measure the average deviation between the predicted value and the true value, while the coefficient of determination reflects the ability of the model to explain the data. After finding that the mean squared error of the preliminary results is high, the system applies bias correction techniques for adjustment. The system identifies the systematic bias in the model through residual analysis and corrects the model. Bias correction improves the prediction accuracy by adjusting the systematic error of the model, while residual analysis is used to identify and correct outliers in the model. After optimization, the system verifies the optimized quantitative results through comparative analysis with known standard samples to evaluate the accuracy and consistency of the quantitative results. The final high-precision quantitative results show the actual concentration of antibiotics in the sample, ensuring the reliability of the quantitative analysis and providing a solid foundation for scientific research and applications.

[0097] It can be seen that according to the physicochemical properties of the target compound, a multi-dimensional gradient elution program is designed to obtain a chromatographic separation spectrum; according to the chromatographic separation spectrum, the target compound is efficiently ionized to obtain an ionization signal; according to the ionization signal, the target ion is fragmented, and through fragment ion screening, a fragment ion spectrum is obtained; according to the fragment ion spectrum, a multivariate data analysis algorithm based on machine learning is adopted, combined with the standard curve of the target compound and the interferent database, for quantitative analysis, and through dynamic interference correction technology, combined with principal component analysis and partial least squares method, a quantitative result is generated, so that the sensitivity, accuracy and real-time performance of the detection can be improved through multi-dimensional gradient elution, dynamic ionization and fragmentation parameter optimization, and analysis means combined with machine learning.

[0098] Another embodiment of the present invention provides a liquid chromatography-tandem mass spectrometry detection system. Refer to Figure 3 , the system may include: A design module 301, configured to design a multi-dimensional gradient elution program according to the physicochemical properties of the target compound, perform liquid chromatography separation on the sample using the elution program, and in combination with a dynamic gradient optimization algorithm, adjust the proportion and flow rate of the mobile phase in real time to obtain a chromatographic separation spectrum; An ionization module 302, configured to perform high-efficiency ionization on the target compound according to the chromatographic separation spectrum, using an ionization efficiency enhancement technique based on electrospray ionization, in combination with a dynamic voltage tuning and solvent-assisted ionization strategy, and dynamically adjust the ion source parameters through an adaptive electric field optimization algorithm to obtain an ionization signal; A screening module 303, configured to fragment the target ions according to the ionization signal, using a multi-stage mass spectrometry fragmentation technique, in combination with a dynamic energy tuning algorithm, and dynamically select characteristic fragment ions through fragment ion screening, in combination with a fragmentation rule library of the target compound, to obtain a fragment ion spectrum; An analysis module 304, configured to perform quantitative analysis according to the fragment ion spectrum, using a multivariate data analysis algorithm based on machine learning, in combination with a standard curve of the target compound and an interferent database, and generate a quantitative result through a dynamic interference correction technique, in combination with principal component analysis and partial least squares method.

[0099] It can be seen that according to the physicochemical properties of the target compound, a multi-dimensional gradient elution program is designed to obtain a chromatographic separation spectrum; according to the chromatographic separation spectrum, high-efficiency ionization of the target compound is performed to obtain an ionization signal; according to the ionization signal, the target ions are fragmented, and through fragment ion screening, a fragment ion spectrum is obtained; according to the fragment ion spectrum, a multivariate data analysis algorithm based on machine learning is used, in combination with a standard curve of the target compound and an interferent database, to perform quantitative analysis, and through a dynamic interference correction technique, in combination with principal component analysis and partial least squares method, a quantitative result is generated, so that the sensitivity, accuracy, and real-time performance of the detection can be improved through multi-dimensional gradient elution, dynamic ionization and fragmentation parameter optimization, and analysis means combined with machine learning.

[0100] The embodiment of the present invention further provides a storage medium, in which a computer program is stored, and wherein the computer program is configured to execute the steps in any one of the above method embodiments when running.

[0101] Specifically, in this embodiment, the above storage medium may be configured to store a computer program for executing the following steps: S201. Design a multi-dimensional gradient elution program according to the physical and chemical properties of the target compound. Use this elution program to perform liquid chromatography separation on the sample. Combine with a dynamic gradient optimization algorithm to adjust the mobile phase ratio and flow rate in real time, and obtain a chromatographic separation spectrum. S202. According to the chromatographic separation spectrum, adopt an ionization efficiency enhancement technique based on electrospray ionization. Combine with a dynamic voltage tuning and solvent-assisted ionization strategy to efficiently ionize the target compound. Through an adaptive electric field optimization algorithm, dynamically adjust the ion source parameters to obtain an ionization signal. S203. According to the ionization signal, adopt a multi-stage mass spectrometry fragmentation technique. Combine with a dynamic energy tuning algorithm to fragment the target ion. Through fragment ion screening, combine with a fragmentation rule library of the target compound to dynamically select characteristic fragment ions and obtain a fragment ion spectrum. S204. According to the fragment ion spectrum, adopt a multi-variable data analysis algorithm based on machine learning. Combine with the standard curve of the target compound and an interferent database to perform quantitative analysis. Through a dynamic interference correction technique, combine with principal component analysis and partial least squares method to generate a quantitative result.

[0102] It can be seen that according to the physical and chemical properties of the target compound, a multi-dimensional gradient elution program is designed to obtain a chromatographic separation spectrum; according to the chromatographic separation spectrum, the target compound is efficiently ionized to obtain an ionization signal; according to the ionization signal, the target ion is fragmented, and through fragment ion screening, a fragment ion spectrum is obtained; according to the fragment ion spectrum, a multi-variable data analysis algorithm based on machine learning is adopted, combined with the standard curve of the target compound and an interferent database, to perform quantitative analysis, and through a dynamic interference correction technique, combined with principal component analysis and partial least squares method, a quantitative result is generated. Thus, it is possible to improve the sensitivity, accuracy, and real-time performance of detection through multi-dimensional gradient elution, dynamic ionization and fragmentation parameter optimization, and analysis means combined with machine learning.

[0103] An embodiment of the present invention also provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0104] Specifically, the above electronic device may further include a transmission device and an input / output device. Among them, the transmission device is connected to the above processor, and the input / output device is connected to the above processor.

[0105] Specifically, in this embodiment, the above processor may be configured to execute the following steps through a computer program: S201. Design a multi-dimensional gradient elution program according to the physical and chemical properties of the target compound. Use the elution program to perform liquid chromatography separation on the sample. Combine with the dynamic gradient optimization algorithm to adjust the mobile phase ratio and flow rate in real time, and obtain a chromatographic separation spectrum. S202. According to the chromatographic separation spectrum, adopt the ionization efficiency enhancement technology based on electrospray ionization, combine the dynamic voltage tuning and solvent-assisted ionization strategies to efficiently ionize the target compound. Through the adaptive electric field optimization algorithm, dynamically adjust the ion source parameters to obtain an ionization signal. S203. According to the ionization signal, adopt the multi-stage mass spectrometry fragmentation technology, combine the dynamic energy tuning algorithm to fragment the target ion. Through fragment ion screening, combine with the fragmentation rule library of the target compound to dynamically select characteristic fragment ions and obtain a fragment ion spectrum. S204. According to the fragment ion spectrum, adopt the multi-variable data analysis algorithm based on machine learning, combine the standard curve of the target compound and the interferent database to perform quantitative analysis. Through the dynamic interference correction technology, combine with principal component analysis and partial least squares method to generate a quantitative result.

[0106] It can be seen that according to the physical and chemical properties of the target compound, a multi-dimensional gradient elution program is designed to obtain a chromatographic separation spectrum; according to the chromatographic separation spectrum, the target compound is efficiently ionized to obtain an ionization signal; according to the ionization signal, the target ion is fragmented, and through fragment ion screening, a fragment ion spectrum is obtained; according to the fragment ion spectrum, a multi-variable data analysis algorithm based on machine learning is adopted, combined with the standard curve of the target compound and the interferent database to perform quantitative analysis, and through the dynamic interference correction technology, combined with principal component analysis and partial least squares method, a quantitative result is generated. Thus, through multi-dimensional gradient elution, dynamic ionization and fragmentation parameter optimization, and combined with machine learning analysis means, the sensitivity, accuracy and real-time performance of detection can be improved.

[0107] The above has described in detail the structure, characteristics and effects of the present invention according to the illustrated embodiments. The above is only the preferred embodiment of the present invention, but the present invention is not limited to the scope defined by the drawings. Any changes made according to the concept of the present invention, or modified into equivalent embodiments with equivalent changes, still within the spirit covered by the specification and drawings, shall be within the protection scope of the present invention.

Claims

1. A liquid chromatography-tandem mass spectrometry detection method, characterized in that The method includes: Designing a multi-dimensional gradient elution program according to the physical and chemical properties of the target compound, performing liquid chromatography separation on the sample using the elution program, and combining with a dynamic gradient optimization algorithm to adjust the mobile phase ratio and flow rate in real time to obtain a chromatographic separation spectrum; According to the chromatographic separation spectrum, adopting an ionization efficiency enhancement technique based on electrospray ionization, combining dynamic voltage tuning and solvent-assisted ionization strategies to efficiently ionize the target compound, and dynamically adjusting the ion source parameters through an adaptive electric field optimization algorithm to obtain an ionization signal; According to the ionization signal, adopting a multi-stage mass spectrometry fragmentation technique, combining a dynamic energy tuning algorithm to fragment the target ion, and dynamically selecting characteristic fragment ions through fragment ion screening and combining with the fragmentation rule library of the target compound to obtain a fragment ion spectrum; According to the fragment ion spectrum, adopting a multivariate data analysis algorithm based on machine learning, combining the standard curve of the target compound and the interferent database to perform quantitative analysis, and generating a quantitative result through a dynamic interference correction technique, combining principal component analysis and partial least squares method.

2. The method according to claim 1, characterized in that The step of designing a multi-dimensional gradient elution program according to the physical and chemical properties of the target compound, performing liquid chromatography separation on the sample using the elution program, and combining with a dynamic gradient optimization algorithm to adjust the mobile phase ratio and flow rate in real time to obtain a chromatographic separation spectrum includes: According to the physical and chemical properties of the target compound, using a prediction model based on molecular descriptors to calculate its retention time and elution conditions, and generating a preliminary elution condition range through a multi-objective optimization algorithm; For the preliminary elution condition range, using a multi-dimensional gradient elution program generation algorithm, combining the three-dimensional parameter space of time-mobile phase ratio-flow rate to design a gradient elution program, and generating a preliminary elution program through dynamic constraint conditions; According to the preliminary elution program, using a dynamic gradient optimization algorithm, combining real-time chromatographic peak shape monitoring data to adjust the mobile phase ratio and flow rate, and optimizing the separation effect through a feedback control mechanism to generate a preliminary high-resolution chromatographic separation spectrum; For the preliminary high-resolution chromatographic separation spectrum, using a peak shape optimization algorithm, combining baseline correction and noise filtering techniques to further improve the spectrum quality, and generating a final high-resolution chromatographic separation spectrum through a resolution evaluation index.

3. The method according to claim 2, wherein The step of adopting an ionization efficiency enhancement technique based on electrospray ionization, combining dynamic voltage tuning and solvent-assisted ionization strategies to efficiently ionize the target compound according to the chromatographic separation spectrum, and dynamically adjusting the ion source parameters through an adaptive electric field optimization algorithm to obtain an ionization signal includes: According to the high-resolution chromatographic separation spectrum, combining the physical and chemical properties of the target compound to initialize the electrospray ionization conditions, and generating preliminary ionization parameters through an ionization efficiency prediction model; For the preliminary ionization parameters, using a dynamic voltage tuning algorithm, combining the solvent-assisted ionization strategy to adjust the spray voltage and solvent ratio in real time, and generating preliminary optimized ionization conditions through an ionization efficiency monitoring technique; For the preliminary optimized ionization conditions, an adaptive electric field optimization algorithm is adopted, combined with ion signal intensity feedback, to dynamically adjust the ion source parameters. Through electric field distribution simulation technology, a preliminary high-quality ionization signal is generated; For the preliminary high-quality ionization signal, signal-to-noise ratio and ion intensity stability evaluation indicators are adopted to optimize the signal quality. Through signal filtering technology, a final high-quality ionization signal is generated.

4. The method according to claim 3, wherein According to the ionization signal, a multi-stage mass spectrometry fragmentation technique is adopted, combined with a dynamic energy tuning algorithm, to fragment the target ions. Through fragment ion screening, combined with the fragmentation rule library of the target compound, characteristic fragment ions are dynamically selected to obtain a fragment ion spectrum, including: According to the high-quality ionization signal, combined with the molecular weight and structural characteristics of the target compound, the fragmentation energy is initialized. Through a fragmentation efficiency prediction model, preliminary fragmentation parameters are generated; For the preliminary fragmentation parameters, a dynamic energy tuning algorithm is adopted, combined with fragment ion intensity feedback, to adjust the collision energy in real time. Through a multi-stage fragmentation strategy, a preliminary fragment ion spectrum is generated; For the preliminary fragment ion spectrum, a screening algorithm based on the fragmentation rule library is adopted, combined with the characteristic fragments of the target compound, to dynamically select characteristic fragment ions. Through fragment ion matching technology, a preliminary highly specific fragment ion spectrum is generated; For the preliminary highly specific fragment ion spectrum, signal-to-noise ratio and fragment ion abundance evaluation indicators are adopted to optimize the spectrum quality. Through spectrum smoothing technology, a final highly specific fragment ion spectrum is generated.

5. The method according to claim 4, characterized in that According to the fragment ion spectrum, a multivariate data analysis algorithm based on machine learning is adopted, combined with the standard curve of the target compound and the interferent database, for quantitative analysis. Through dynamic interference correction technology, combined with principal component analysis and partial least squares method, a quantitative result is generated, including: According to the highly specific fragment ion spectrum, data preprocessing technology is adopted, combined with a feature extraction algorithm, to generate a preliminary quantitative feature data set; For the preliminary quantitative feature data set, dynamic interference correction technology is adopted, combined with the interferent database, to identify and correct the interferent signals. Through principal component analysis and partial least squares method, a preliminary corrected data set is generated; For the preliminary corrected data set, a multivariate data analysis algorithm based on machine learning is adopted, combined with the standard curve of the target compound, for quantitative analysis. Through model optimization technology, a preliminary quantitative result is generated; For the preliminary quantitative result, accuracy evaluation indicators are adopted to optimize the quantitative result. Through an error correction mechanism, a final high-precision quantitative result is generated.

6. A liquid chromatography-tandem mass spectrometry detection system, characterized in that, The system includes: A design module, used to design a multi-dimensional gradient elution program according to the physicochemical properties of the target compound, use the elution program to perform liquid chromatography separation on the sample, and combine a dynamic gradient optimization algorithm to adjust the mobile phase ratio and flow rate in real time to obtain a chromatographic separation spectrum; An ionization module, used to adopt an ionization efficiency enhancement technology based on electrospray ionization according to the chromatographic separation spectrum, combine a dynamic voltage tuning and solvent-assisted ionization strategy to efficiently ionize the target compound, and dynamically adjust the ion source parameters through an adaptive electric field optimization algorithm to obtain an ionization signal; A screening module, which is used to fragment target ions according to ionization signals, adopt multi-stage mass spectrometry fragmentation technology, and combine with a dynamic energy tuning algorithm. Through fragment ion screening, and by combining with a fragmentation rule library of target compounds, it dynamically selects characteristic fragment ions to obtain a fragment ion spectrum diagram. An analysis module, which is used to perform quantitative analysis according to the fragment ion spectrum diagram, adopt a multivariate data analysis algorithm based on machine learning, and combine with a standard curve of target compounds and an interferent database. Through dynamic interference correction technology, and by combining with principal component analysis and partial least squares method, it generates a quantitative result.

7. The system according to claim 6, wherein The design module is specifically used for: According to the physical and chemical properties of target compounds, adopt a prediction model based on molecular descriptors to calculate their retention times and elution conditions, and generate a preliminary elution condition range through a multi-objective optimization algorithm. For the preliminary elution condition range, adopt a multi-dimensional gradient elution program generation algorithm, combine with a three-dimensional parameter space of time-mobile phase ratio-flow rate, design a gradient elution program, and generate a preliminary elution program through dynamic constraint conditions. According to the preliminary elution program, adopt a dynamic gradient optimization algorithm, combine with real-time chromatographic peak shape monitoring data, adjust the mobile phase ratio and flow rate, and optimize the separation effect through a feedback control mechanism to generate a preliminary high-resolution chromatographic separation spectrum diagram. For the preliminary high-resolution chromatographic separation spectrum diagram, adopt a peak shape optimization algorithm, combine with baseline correction and noise filtering technology, further improve the spectrum quality, and generate a final high-resolution chromatographic separation spectrum diagram through a resolution evaluation index.

8. The system according to claim 7, wherein The ionization module is specifically used for: According to the high-resolution chromatographic separation spectrum diagram, combine with the physical and chemical properties of target compounds, initialize the electrospray ionization conditions, and generate preliminary ionization parameters through an ionization efficiency prediction model. For the preliminary ionization parameters, adopt a dynamic voltage tuning algorithm, combine with a solvent-assisted ionization strategy, adjust the spray voltage and solvent ratio in real time, and generate preliminary optimized ionization conditions through an ionization efficiency monitoring technology. For the preliminary optimized ionization conditions, adopt an adaptive electric field optimization algorithm, combine with ion signal intensity feedback, dynamically adjust the ion source parameters, and generate preliminary high-quality ionization signals through an electric field distribution simulation technology. For the preliminary high-quality ionization signals, adopt signal-to-noise ratio and ion intensity stability evaluation indexes to optimize the signal quality, and generate final high-quality ionization signals through a signal filtering technology.

9. A storage medium, characterized in that, The computer program is stored in the storage medium, wherein the computer program is set to execute the method according to any one of claims 1-5 when running.

10. An electronic device, comprising a memory and a processor, characterized in that, The computer program is stored in the memory, and the processor is set to run the computer program to execute the method according to any one of claims 1-5.

Images