An intelligent analysis system and method for identifying dimer isomers in β-lactam antibiotics

Abstract

The present invention discloses a kind of intelligent analysis system for identifying dimer isomers in beta-lactam antibiotics, each dimer isomer and its potential characteristic fragment data of theoretical calculation are obtained by potential characteristic fragment data collection subsystem, including the molecular weight of potential characteristic fragment in positive ion mode; the potential characteristic fragment data of each dimer isomer are screened by unique characteristic fragment data screening subsystem, and its corresponding unique characteristic fragment data are obtained; the mass spectral image obtained by detection is selected by target mass spectral image under target molecular weight by mass spectral image recognition subsystem, and its mass spectral image data are extracted, including the mass-to-charge ratio corresponding to fragment ion peak; each unique characteristic fragment data and each mass spectral image recognition data are compared by comparison subsystem, and the molecular weight and mass-to-charge ratio therein are obtained according to a preset fragment matching strategy. The present invention can replace detection personnel to carry out effective screening of dimer isomer impurities.

Description

Technical Field

[0001] The present invention relates to the field of drug analysis and quality control, in particular to an intelligent analysis system and method for dimer isomers in beta-lactam antibiotics. Background Art

[0002] β-lactam antibiotics (β-lactams) refer to a large class of antibiotics containing a β-lactam ring in their chemical structure. They are widely used clinically to treat sepsis, pneumonia, lung abscesses, secondary infections caused by sensitive bacteria, chronic respiratory diseases, refractory cystitis, pyelonephritis, peritonitis, and gynecological adnexitis. Although this product has a broad antimicrobial spectrum and strong antimicrobial activity, it is extremely unstable and easily degrades in the presence of moisture and heat, producing highly allergenic polymer impurities that pose a serious threat to patient safety.

[0003] Currently, most analyses of allergenic polymer impurities in β-lactam antibiotics rely on instrumental quantitative analysis. The structures of these polymer impurities remain unclear, and the number of polymer impurities detected is limited. Several studies have found that isomeric impurities in polymer impurities are closely associated with their toxic effects; the European Pharmacopoeia EP 8.0 and the United States Pharmacopoeia USP 37 / NF32 include several dimer isomers, highlighting the importance of isomeric identification. However, the structural similarity of isomeric impurities to the main drug components presents a challenge in impurity identification. Therefore, it is necessary to develop new approaches to investigate potential polymerization sites of polymer impurities and characterize the structural characteristics of isomers to facilitate impurity control in β-lactam antibiotics. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent analysis system and method for identifying dimer isomers in β-lactam antibiotics, so as to solve the problem of difficulty in impurity identification due to the close structure of dimer isomer impurities and the main components of drugs raised in the background art.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An intelligent analysis system for identifying dimer isomers in β-lactam antibiotics, comprising:

[0007] The potential characteristic fragment data collection subsystem is used to intelligently read dimer isomers and their potential characteristic fragments using theoretical cleavage fragments of dimer impurities, and obtain theoretically calculated potential characteristic fragment data of each dimer isomer, including the molecular weight of the potential characteristic fragment in positive ion mode;

[0008] The unique characteristic fragment data screening subsystem is used to screen the potential characteristic fragment data of each dimer isomer to obtain its corresponding unique characteristic fragment data, including the molecular weight of the unique characteristic fragment in positive ion mode;

[0009] The mass spectrum image recognition subsystem is used to select the target mass spectrum image at the target molecular weight from the mass spectrum image obtained by the actual dimer impurity detection, and extract its mass spectrum image data, including the mass-to-charge ratio corresponding to the fragment ion peak;

[0010] The comparison subsystem is used to compare the theoretically calculated unique characteristic fragment data with the actual mass spectrum image recognition data of the dimer impurity, and obtain the matching results for the molecular weight and mass-to-charge ratio according to the preset fragment matching strategy.

[0011] Preferably, the potential feature segment data collection subsystem includes:

[0012] A first data reading module is used to read theoretical cleavage fragments of dimer impurities;

[0013] The feature collection module is used to list the polymerization modes based on each theoretical cleavage fragment and its different potential polymerization sites, obtain several different dimer isomers and collect relevant feature data;

[0014] a calculation module for performing relevant characteristic data processing on each set of potential characteristic fragments of each dimer isomer, wherein the potential characteristic fragments are fragments obtained by cutting the dimer isomer according to the cleavage rules of carbapenem compounds while maintaining the integrity of the amide or ester bond in the dimer connecting the two molecules;

[0015] The first data output module is used to output data of each dimer isomer and its potential characteristic fragments, wherein each potential characteristic fragment data includes the molecular weight of each potential characteristic fragment in a positive ion mode.

[0016] Preferably, the unique feature segment data screening subsystem includes:

[0017] A second data reading module is used to read potential characteristic fragment data of each dimer isomer;

[0018] The feature screening module is used to eliminate the potential feature fragments shared by other dimer isomers from the potential feature fragment data to obtain unique feature fragment data for each dimer isomer;

[0019] The second data output module is used to output the unique characteristic fragment data of each dimer isomer, including the molecular weight of the unique characteristic fragment in the positive ion mode.

[0020] Preferably, the mass spectrum image recognition subsystem includes:

[0021] A third data reading module is used to read the mass spectrum image actually detected;

[0022] The image recognition and conversion module is used to determine the RT range of the dimer impurity in the total ion current map based on the target molecular weight, obtain the mass spectrum image within the RT range as the target mass spectrum image, and extract mass spectrum image recognition data for any target mass spectrum image, including extracting mass-to-charge ratio and mass spectrum peak intensity in positive ion mode;

[0023] The third data output module is used to output each mass spectrum image recognition data within the target RT range.

[0024] Preferably, in the image recognition module, the number of target mass spectrometry images is no less than 3.

[0025] Preferably, the comparison subsystem includes:

[0026] A fourth data reading module is used to read each unique characteristic fragment data and each mass spectrum image recognition data;

[0027] The accuracy control module is used to set the fragment matching strategy and its matching accuracy, wherein the fragment matching strategy includes a first matching strategy and a second matching strategy; in the first matching strategy, when the absolute difference between the molecular weight and the mass-to-charge ratio does not exceed a preset value, the corresponding unique feature fragment data is determined to match the mass spectrum image recognition data; in the second matching strategy, when the same unique feature fragment data matches no less than two mass spectrum image recognition data in the same RT interval, the corresponding dimer isomer is determined to be consistent with the dimer impurity in the RT interval.

[0028] A comparison module is used to match the molecular weight of the unique characteristic fragment data and the mass-to-charge ratio of the mass spectrum image recognition data according to a preset fragment matching strategy;

[0029] The fourth output module is used to output the matching result.

[0030] Preferably, the comparison subsystem further includes a noise reduction module for removing background impurity peaks in the mass spectrum image recognition data.

[0031] Preferably, the intelligent analysis system further includes a network visualization module for visualizing the matching results.

[0032] An intelligent analytical method for identifying dimer isomers in β-lactam antibiotics, the method comprising the following steps:

[0033] Implementation prerequisites: Conduct actual drug testing and theoretical simulations to obtain mass spectrometry images and theoretical fragment sets of dimer impurities;

[0034] For each theoretical cleavage fragment, the polymerization mode is listed according to the potential polymerization site to obtain a number of dimer isomers, all potential characteristic fragments of each dimer isomer are extracted and their molecular weight is calculated to obtain the corresponding potential characteristic fragment data;

[0035] screening the potential characteristic fragment data for unique characteristic fragment data of each dimer isomer;

[0036] The mass spectrometry images obtained by detection are screened using the target molecular weight and the RT range in the total ion current graph in which it is located. Several mass spectrometry images within the RT range are selected as target mass spectrometry images. The mass-to-charge ratio corresponding to the fragment ion peak of each target mass spectrometry image obtained by screening is extracted to obtain the corresponding mass spectrometry image recognition data;

[0037] The mass spectrum image recognition data and the unique characteristic fragment data are matched according to the preset fragment matching strategy to obtain the matching result.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] The intelligent analysis system of the present invention constructs potential dimer isomer impurity configurations through theoretical cleavage fragments of dimer impurities. By matching the unique characteristic fragments of the isomer impurity configurations with mass spectrometry data, it can effectively screen out the configurations of isomer impurities, providing a new path for the identification of β-lactam antibiotic polymer impurities. The intelligent analysis system of the present invention can replace the manual reading and analysis of instrument data by testers to perform intelligent and effective identification of dimer isomer impurities. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 Schematic diagram of the framework structure of the intelligent analysis system.

[0041] Figure 2 for Figure 1 Schematic diagram of the specific structure of the potential feature fragment data collection subsystem.

[0042] Figure 3 for Figure 1 Schematic diagram of the specific structure of the unique feature fragment data screening subsystem.

[0043] Figure 4 for Figure 1 Schematic diagram of the specific structure of the mass spectrometry image recognition subsystem.

[0044] Figure 5 for Figure 1 Schematic diagram of the specific structure of the comparison subsystem.

[0045] Figure 6 This is an analytical diagram of the implementation method of the network visualization module.

[0046] Figure 7 Schematic diagram of the implementation conditions and materials of the intelligent analysis method.

[0047] Figure 8 This is an example diagram of the relationship between potential feature fragment data and unique feature fragment data.

[0048] Figure 9 This is an example diagram of the correspondence between the mass spectrum image actually detected in the mass spectrum image recognition subsystem and its mass spectrum image recognition data.

[0049] Figure 10 This is an example diagram of the matching between unique feature fragment data and its mass spectrum image recognition data in the alignment subsystem.

[0050] Figure 11 This is the visualization output result of the network visualization module under different option settings. DETAILED DESCRIPTION

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0052] Reference Figure 1 As shown, an intelligent analysis system for dimer isomers in β-lactam antibiotics consists of a potential feature fragment data collection subsystem, a unique feature fragment data screening subsystem, a mass spectrometry image recognition subsystem, a comparison subsystem, and a network visualization module.

[0053] Reference Figure 2 The potential feature segment data collection subsystem includes a first data reading module, a feature acquisition module, a calculation module, and a first data output module.

[0054] The first data reading module is used to read drug information, drug sample source information, and a theoretical cleavage fragment set, where the theoretical cleavage fragment set is all possible dimer impurity cleavage fragments.

[0055] The feature acquisition module is used to arrange each theoretically understood fragment according to different potential polymerization sites to obtain relevant feature data of all possible dimer isomers.

[0056] The calculation module is used to develop and run an acquisition code for each dimer isomer to obtain its corresponding potential characteristic fragment data, including molecular weight data of each potential characteristic fragment that may be generated by fragmentation in a hydrogen-containing positive ion mode.

[0057] In the feature acquisition module of the present invention, since the theoretical fragments of dimer impurities all have multiple potential polymerization sites, possible dimer isomers can be obtained based on the possible polymerization sites between monomers. For example, in penem-type drugs, dimer isomers may be formed by the connection of a carboxyl group with an amino or imino group, or a carboxyl group with a hydroxyl group. For penicillin drugs, dimer formation may be caused by the polymerization of a carboxyl group with an amino or imino group.

[0058] In the calculation module of the present invention, while keeping the amide or ester bond in the dimer connecting the two molecules intact, possible dimer isomers are cut according to the cleavage rules of carbapenem compounds to obtain fragments that may be generated by the breakage of the dimer isomer in a hydrogen-containing positive ion mode. The fragments generated by the cleavage of the same dimer isomer constitute a potential characteristic fragment set, that is, a cleavage sample. The potential characteristic fragment set under each dimer configuration will contain fragments with potential structural differences from other dimer configurations.

[0059] The first data output module is used to output the potential characteristic fragment data of each dimer isomer, including fragment document code, isomer and its number, drug name, cleavage sample and its molecular weight (positive ion).

[0060] Reference Figure 3 As shown, the unique feature segment data screening subsystem includes a second data reading module, a feature screening module, and a second data output module.

[0061] The second data reading module is used to read the potential characteristic fragment data of each dimer isomer, including the fragment file code, isomer and its number, drug name, cleavage sample and its molecular weight.

[0062] The feature screening module is used to eliminate the potential feature fragments shared by other dimer isomers from the potential feature fragment data to obtain unique feature fragment data for each dimer isomer.

[0063] The second data output module is used to output the unique characteristic fragment data (UNIQUE profile) of each dimer isomer, including the unique characteristic fragment file code, isomer number, drug information, cleavage sample and its molecular weight (positive ion).

[0064] Reference Figure 4 The mass spectrometry image recognition subsystem includes a third data reading module, an image recognition module, and a third data output module.

[0065] The third data reading module is used to read mass spectrometry image data, including drug name, sample conditions, mass spectrometry instrument name and model, operation mode, and actually detected LC-MS / MS mass spectrometry image.

[0066] The image data recognition and conversion module is used to identify LC-MS / MS mass spectrometry images and, based on the target molecular weight and the RT range in the total ion current map in which it is located, select several mass spectrometry images within the RT range as target mass spectrometry images. The module then extracts the mass-to-charge ratio and mass spectrometry peak intensity of any target mass spectrometry image in the positive ion mode to obtain the corresponding mass spectrometry image recognition data (MSDATA).

[0067] The total ion current diagram is a commonly used graph in mass spectrometry analysis. It shows the change in the signal intensity of all ions in the sample over time. It is an experimental data graph obtained from the actual detected LC-MS / MS mass spectrum image.

[0068] The third data output module is used to output each mass spectrometry image recognition data, including image data document code, drug name, sample conditions, target molecular weight (positive ion), RT value and RT range, mass spectrometry instrument name and model.

[0069] Reference Figure 5 The comparison subsystem includes a fourth data reading module, an accuracy control module, a comparison module, and a fourth data output module.

[0070] The fourth data reading module is used to read the unique feature fragment data output by the unique feature fragment data screening subsystem and the mass spectrum image recognition data output by the mass spectrum image recognition subsystem.

[0071] The accuracy control module is used to set a fragment matching strategy and its matching accuracy. The fragment matching strategy includes a first matching strategy and a second matching strategy.

[0072] First matching strategy: If the absolute difference between the molecular weight (positive ion) of the unique characteristic fragment in the unique characteristic fragment data and the mass-to-charge ratio in the mass spectrum image recognition data does not exceed a preset value, then the dimer isomer corresponding to the unique characteristic fragment data is matched with the mass spectrum image recognition data. The specific representation formula of the absolute difference is:

[0073]

[0074] Wherein, D represents the difference value, m represents the charge-to-mass ratio in the mass spectrometry image recognition data under positive ion mode, and t represents the charge-to-mass ratio of the unique characteristic fragment in the theoretical unique characteristic fragment data under positive ion mode.

[0075] The second matching strategy: if the same unique characteristic fragment data matches no less than two mass spectrum image recognition data in the same RT interval, then the substance in the interval is considered to correspond to the dimer isomer of the unique characteristic fragment data.

[0076] The second matching strategy is completed on the basis of the first matching strategy. The first matching strategy is used to compare and match the unique characteristic fragment data of the dimer isomer with the mass spectrometry image recognition data of the same target molecular weight in each RT interval. After the matching is completed, the second matching strategy is executed based on the matching.

[0077] In the present invention, the preset value in the first matching strategy is the matching accuracy, which is generally between 5-20 ppm. The specific value of the matching accuracy can be set by those skilled in the art according to actual conditions.

[0078] The comparison module is used to match the theoretical characteristic fragment data of the dimer isomer with the actual dimer impurity mass spectrum image recognition data based on a preset fragment matching strategy to obtain a matching result.

[0079] The fourth data output module is used to output the matching results, including the file names of the unique feature fragment data and mass spectrometry image recognition data that are successfully matched under the first matching strategy and the identical values ​​between the two, and / or the unique feature fragment data of the dimer isomer that are successfully matched under the second matching strategy, etc.

[0080] In the present invention, the matching accuracy is set by setting an accuracy control module, thereby improving the accuracy and specificity of β-lactam antibiotic impurity risk control.

[0081] Furthermore, the comparison subsystem also includes a noise reduction module for filtering out background impurity peaks with no detection significance in the mass spectrum image recognition data.

[0082] The Network Visualization module visualizes data based on the matching results output by the alignment subsystem, displaying relevant data and relationship networks for the unique characteristic fragments of the matched dimer isomers, including the corresponding file name, mass spectrum peak and RT value, alignment value, molecular weight in positive ion mode, cleavage sample, and fragment structure formula. The "Visualization Options" function within the module allows users to select the display of relationship data, such as unique data visualization, the correspondence between Excel data and unique relationship data, mass spectrum data, and the unique characteristic fragment structure formula.

[0083] The intelligent analysis system of the present invention comprises a potential characteristic fragment data collection subsystem that performs theoretical simulations of theoretical dimer isomers and their potential characteristic fragments through code development and intelligent data processing, and a unique characteristic fragment data screening subsystem that screens dimer isomers for unique characteristic fragment data; a mass spectrum image recognition subsystem extracts mass spectrum image recognition data from mass spectrum image information of actually detected impurities, and a comparison subsystem achieves precise pairing of the mass spectrum image recognition data with the unique characteristic fragment data, thereby realizing a multi-dimensional presentation of the relevant data relationship network.

[0084] Based on the above intelligent analysis system, the present invention also provides an intelligent analysis method, which will be specifically described below using drug a as an example.

[0085] Preconditions for implementation: Refer to Figure 7 By performing actual detection and theoretical simulation on drug a, mass spectrometry images and all possible cleavage fragments of the dimer impurity are obtained, and all possible cleavage fragments constitute a theoretical cleavage fragment set.

[0086] Step 1: For each theoretical cleavage fragment, the polymerization mode is listed according to the potential polymerization site to obtain all possible dimer forms. Each dimer is cut according to the cleavage rules of carbapenem compounds while keeping the amide or ester bond in the dimer connecting the two molecules intact to obtain the corresponding potential characteristic fragment set, and the molecular weight of each potential characteristic fragment in positive ion mode is calculated to obtain the corresponding potential characteristic fragment data.

[0087] In step 1 of the present invention, for each theoretical cleavage fragment, different dimers can be obtained after polymerization according to its different potential polymerization sites. Different dimers of different theoretical cleavage fragments are cut separately to obtain their corresponding potential characteristic fragment data, including the molecular weight of each potential characteristic fragment and the corresponding dimer configuration, etc. Figure 8 As shown in (a).

[0088] Step 2: Screen the potential characteristic fragment data of each dimer isomer to obtain the unique characteristic fragment data of each dimer isomer, such as Figure 8 (b) shown.

[0089] In step 2 of the present invention, the potential characteristic fragments shared by other dimer isomers are eliminated from the potential characteristic fragment set of each dimer isomer, and the potential characteristic fragments that are not eliminated are used as unique characteristic fragments of the dimer isomer.

[0090] Step 3: For the mass spectrum images obtained from the actual detection of drug a, the target molecular weight and the RT range in the total ion current graph are used to screen the mass spectrum images, and at least three mass spectrum images within the RT range are selected as target mass spectrum images. Image data extraction is performed on each target mass spectrum image obtained by screening, as shown in FIG. Figure 9 As shown in (b), the image data includes the mass-to-charge ratio and mass spectrum peak intensity in positive ion mode.

[0091] In step 3 of the present invention, the target molecular weight is the molecular weight of the dimer impurity. In this example, the molecular weight of the dimer impurity of Drug A is 701. The molecular weight of the dimer impurity can be calculated based on the structure of Drug A, which is common knowledge in the art. The target mass spectrometry image and the mass spectrometry image recognition data are in a one-to-one correspondence.

[0092] Step 4: Match the mass spectrum image recognition data of each actual dimer impurity with the theoretical unique characteristic fragment data according to the first matching strategy to obtain a matching result.

[0093] Figure 10 As shown, in step 4 of this embodiment, the preset value in the first matching strategy is set to 10ppm; the unique characteristic fragment data with a molecular weight of 701 is compared with each mass spectrum image recognition data to obtain a matching relationship between each target mass spectrum image with a mass-to-charge ratio of 701m / z in the positive ion mode and each unique characteristic fragment data, and the matching relationship is recorded in Excel, wherein column A represents the RT interval where the successfully matched target mass spectrum image is located, columns B and C are used to characterize the successfully matched target mass spectrum image and the unique characteristic fragment data, respectively, and columns D and E represent the mass-to-charge ratio of the same target mass spectrum image in column B and the fragment molecular weight of the same unique characteristic fragment data in column C, respectively, and the absolute difference values ​​of the mass-to-charge ratio and the fragment molecular weight are both less than 10ppm.

[0094] Step 5: Visualize the matching results according to different options, such as Figure 11 shown.

[0095] The present invention screens out the unique characteristic fragment data of each dimer isomer theoretically calculated from the actual detected impurities and their mass spectrometry image information through code development and data intelligent processing, thereby achieving accurate pairing of mass spectrometry detection data with potential dimer isomer characteristic data and realizing multi-dimensional display of the relevant data relationship network. The present invention provides a new path for the identification of β-lactam antibiotic polymer impurities and also improves the accuracy and specificity of β-lactam antibiotic impurity risk control. The technical solution of the present invention is applicable to other types of drugs besides β-lactam antibiotics and can be promoted and used in the drug analysis and quality control industries.

Claims

1. An intelligent analysis system for identifying dimer isomers in β-lactam antibiotics, characterized in that: include: The potential characteristic fragment data collection subsystem is used to intelligently read dimer isomers and their potential characteristic fragments using theoretical cleavage fragments of dimer impurities, and obtain theoretically calculated potential characteristic fragment data of each dimer isomer, including the molecular weight of the potential characteristic fragment in positive ion mode; The unique characteristic fragment data screening subsystem is used to screen the potential characteristic fragment data of each dimer isomer to obtain its corresponding unique characteristic fragment data, including the molecular weight of the unique characteristic fragment in positive ion mode; The mass spectrum image recognition subsystem is used to select the target mass spectrum image at the target molecular weight from the mass spectrum image obtained by the actual dimer impurity detection, and extract its mass spectrum image data, including the mass-to-charge ratio corresponding to the fragment ion peak; The comparison subsystem is used to compare the theoretically calculated unique characteristic fragment data with the actual mass spectrum image recognition data of the dimer impurity, and obtain the matching results according to the preset fragment matching strategy for the molecular weight and mass-to-charge ratio; The potential feature segment data collection subsystem includes: A first data reading module is used to read theoretical cleavage fragments of dimer impurities; The feature collection module is used to list the polymerization modes based on each theoretical cleavage fragment and its different potential polymerization sites, obtain several different dimer isomers and collect relevant feature data; a calculation module for performing relevant characteristic data processing on each set of potential characteristic fragments of each dimer isomer, wherein the potential characteristic fragments are fragments obtained by cutting the dimer isomer according to the cleavage rules of carbapenem compounds while maintaining the integrity of the amide or ester bond in the dimer connecting the two molecules; A first data output module is used to output data of each dimer isomer and its potential characteristic fragments, wherein each potential characteristic fragment data includes a molecular weight of each potential characteristic fragment in a positive ion mode; The comparison subsystem includes: A fourth data reading module is used to read each unique characteristic fragment data and each mass spectrum image recognition data; an accuracy control module for setting a fragment matching strategy and its matching accuracy, wherein the fragment matching strategy includes a first matching strategy and a second matching strategy; in the first matching strategy, when the absolute difference between the molecular weight and the mass-to-charge ratio does not exceed a preset value, the corresponding unique characteristic fragment data is determined to match the mass spectrum image recognition data; in the second matching strategy, when the same unique characteristic fragment data matches no less than two mass spectrum image recognition data in the same RT interval, the corresponding dimer isomer is determined to be consistent with the dimer impurity in the RT interval; A comparison module is used to match the molecular weight of the unique characteristic fragment data and the mass-to-charge ratio of the mass spectrum image recognition data according to a preset fragment matching strategy; The fourth output module is used to output the matching result.

2. The intelligent analysis system for identifying dimer isomers in β-lactam antibiotics according to claim 1, wherein the unique feature fragment data screening subsystem comprises: A second data reading module is used to read potential characteristic fragment data of each dimer isomer; The feature screening module is used to eliminate the potential feature fragments shared by other dimer isomers from the potential feature fragment data to obtain unique feature fragment data for each dimer isomer; The second data output module is used to output the unique characteristic fragment data of each dimer isomer, including the molecular weight of the unique characteristic fragment in the positive ion mode.

3. The intelligent analysis system for identifying dimer isomers in β-lactam antibiotics according to claim 1, characterized in that: The mass spectrum image recognition subsystem includes: A third data reading module is used to read the mass spectrum image actually detected; The image recognition and conversion module is used to determine the RT range of the dimer impurity in the total ion current map based on the target molecular weight, obtain the mass spectrum image within the RT range as the target mass spectrum image, and extract mass spectrum image recognition data for any target mass spectrum image, including extracting mass-to-charge ratio and mass spectrum peak intensity in positive ion mode; The third data output module is used to output each mass spectrum image recognition data within the target RT range.

4. The intelligent analysis system for identifying dimer isomers in β-lactam antibiotics according to claim 3, characterized in that: In the image recognition module, the number of target mass spectrum images is no less than 3.

5. The intelligent analysis system for identifying dimer isomers in β-lactam antibiotics according to claim 1, characterized in that: The comparison subsystem also includes a noise reduction module for removing background impurity peaks in the mass spectrum image recognition data.

6. The intelligent analysis system for identifying dimer isomers in β-lactam antibiotics according to claim 1, characterized in that: The intelligent analysis system further includes a network visualization module for visualizing the matching results.

7. An intelligent analysis method for identifying dimer isomers in β-lactam antibiotics, using the intelligent analysis system according to any one of claims 1 to 6, characterized in that: The method comprises the following steps: implementing preconditions: performing actual testing and theoretical simulation on the drug to obtain a mass spectrum image and a set of theoretical cleavage fragments of the dimer impurity; For each theoretical cleavage fragment, the polymerization mode is listed according to the potential polymerization site to obtain a number of dimer isomers, all potential characteristic fragments of each dimer isomer are extracted and their molecular weight is calculated to obtain the corresponding potential characteristic fragment data; screening the potential characteristic fragment data for unique characteristic fragment data of each dimer isomer; The mass spectrometry images obtained by detection are screened using the target molecular weight and the RT range in the total ion current graph in which it is located. Several mass spectrometry images within the RT range are selected as target mass spectrometry images. The mass-to-charge ratio corresponding to the fragment ion peak of each target mass spectrometry image obtained by screening is extracted to obtain the corresponding mass spectrometry image recognition data; The mass spectrum image recognition data and the unique characteristic fragment data are matched according to the preset fragment matching strategy to obtain the matching result.

Images