An automatic processing method compatible with multi-source mass spectrum data

Abstract

The application belongs to the technical field of mass spectrometry detection, and provides an automatic processing method compatible with multi-source mass spectrometry data. The method acquires mass spectrometry data and calls corresponding parameter information, extracts ion flow and locates candidate signal regions, preliminarily judges based on a judgment condition, and when the judgment is not clear or there is a risk of interference, further analyzes peak type characteristics to identify abnormal signals, finally outputs a processing conclusion by comprehensively judging and identifying results, and saves intermediate data. The scheme provided by the application realizes optimization for specific detection scenarios, avoids problems such as inaccurate integration and peak recognition deviation caused by default processing of general software, significantly improves the accuracy, consistency and automation level of mass spectrometry data processing, reduces the workload of manual review and the dependence on experience judgment, and meets the actual application requirements of the detection industry for high efficiency, stability and automation.

Description

Technical Field

[0001] This invention belongs to the field of mass spectrometry detection technology, specifically relating to an automatic processing method compatible with multi-source mass spectrometry data. Background Technology

[0002] Mass spectrometry is widely used in scientific research, medicine, environment, and chemical engineering. Currently, mass spectrometry data processing typically relies on the software provided by the instrument manufacturer. However, such software is mostly general-purpose tools for multiple applications, with relatively universal data processing logic. It is largely based on fixed parameter configurations and single judgment logic, lacking effective identification mechanisms for complex situations where the judgment results are unclear or there is a risk of interference. In many cases, the processing accuracy is still insufficient, making it difficult to fully meet the actual needs of the detection industry for specific target substances, specific methods, and specific judgment criteria.

[0003] Furthermore, in actual testing, test results often cannot be determined solely based on the automatic output of the software. They also require comprehensive analysis in conjunction with various other information. If the default processing results of general software are used directly, deviations and misjudgments are likely to occur. Therefore, existing testing solutions still rely heavily on experience-based judgment and manual verification. Summary of the Invention

[0004] To overcome the shortcomings of the prior art, this invention proposes an automatic processing method compatible with multi-source mass spectrometry data, the method comprising:

[0005] Acquire mass spectrometry data to be processed, determine a detection method based on the mass spectrometry data, and call parameter information based on the mass spectrometry data and / or the detection method; the parameter information is determined based on preset detection logic and / or other mass spectrometry data belonging to the same batch as the mass spectrometry data, and the parameter information includes at least characteristic ion information and judgment condition information; Based on the characteristic ion information, ion flow is extracted from the mass spectrometry data, and candidate signal regions are located within a preset retention time range; The candidate signal region is determined based on the determination condition information. When the determination result is unclear or the determination result indicates that there is a risk of interference, the peak shape characteristics of the candidate signal region corresponding to the mass spectrometry data are analyzed to identify abnormal signals. Based on the combined judgment results and the identified abnormal signals, a processing conclusion for the mass spectrometry data is output, and intermediate data during the detection process is saved.

[0006] The above-mentioned technical solution has at least the following beneficial effects: The method has excellent adaptability to multi-source mass spectrometry data. By dynamically calling the corresponding parameter information based on mass spectrometry data and / or detection methods, it can be customized for different detection objects and different detection methods. At the same time, by performing regular judgment on candidate signal regions based on judgment condition information, it avoids the deviation that may occur in the default integration and default peak recognition results of general software in specific scenarios, effectively improving the accuracy of target object recognition and reducing the probability of problems such as inaccurate integration, peak recognition deviation, and misjudgment of interference peaks. When the results of the rule-based judgment are unclear or there is a risk of interference, the peak characteristics of the candidate signal region are further analyzed. The traditional judgment based solely on retention time and intensity is extended to the recognition of the peak shape itself. This can effectively handle complex situations such as weak target peak signals, strong background interference, irregular peak shapes, and difficulty in separating overlapping peaks, thus improving the recognition capability under weak signal peaks and ambiguous boundary conditions. By combining the results of rule-based judgment and peak shape feature analysis, the system automatically outputs processing conclusions and saves intermediate data for traceability, which significantly improves the level of automation, reduces the workload of manual intervention, and improves the consistency of results between different operators and different batches. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 This is a schematic diagram of the method flow for the automatic processing method compatible with multi-source mass spectrometry data provided in Example 1; Figures 2(a)-2(d) are all schematic diagrams containing abnormal peak shapes. Figure 2(a) is a schematic diagram containing an overload peak, Figure 2(b) is a schematic diagram containing a tailing peak, Figure 2(c) is a schematic diagram containing a double peak, and Figure 2(d) is a schematic diagram containing a shoulder peak. Detailed Implementation

[0009] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0010] Various embodiments of the invention will be described more fully below. The invention may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of the invention to the specific embodiments disclosed herein, but rather the invention should be understood to cover all modifications, equivalents, and / or alternatives falling within the spirit and scope of the various embodiments of the invention.

[0011] Example 1 This embodiment proposes an automated processing method compatible with multi-source mass spectrometry data. By calling corresponding parameters according to different detection objects or methods, the method solves the problem of insufficient specificity of general-purpose software in specific detection scenarios, improving the flexibility of parameter configuration. Through the establishment of dedicated rules for the actual judgment needs of the detection industry, it can handle complex peak shapes, irregular peaks, weak signal peaks, and ambiguous boundary conditions, effectively avoiding problems such as inaccurate integration, peak identification deviation, and misjudgment of interference peaks. This improves the identification accuracy in complex situations and reduces reliance on human experience. Please refer to [link to previous document]. Figure 1 The methods specifically include: S100: Acquire the mass spectrometry data to be processed, determine the detection method based on the mass spectrometry data, and call the parameter information based on the mass spectrometry data and / or the detection method.

[0012] In this embodiment, the parameter information is determined based on a preset detection logic and / or other mass spectrometry data belonging to the same batch as the mass spectrometry data. The parameter information includes at least characteristic ion information and judgment condition information.

[0013] Specifically, when acquiring mass spectrometry data to be processed, step S100 can directly read the original mass spectrometry data from the instrument manufacturer, or acquire standard format data such as mzXML and mzML obtained by converting the mass spectrometry data using an open-source conversion tool.

[0014] Specifically, the method for dividing multiple mass spectrometry data to be processed into sample data in the same batch includes determining the mass spectrometry data in the same batch as blank sample data, standard sample data or sample data to be tested based on the sample name, sample sequence information, method report, file path, metadata fields, preset keywords and / or ion response characteristics corresponding to the mass spectrometry data. In this embodiment, the preset rules for dividing multiple mass spectrometry data to be processed into sample data in the same batch include one or more of the following: sample sequence continuity, data generation time, file directory, detection method identifier, standard sample location, sample name keywords and / or quantity threshold. For example, the method proposed in this embodiment can sequentially acquire multiple mass spectrometry data during a continuous detection process, and when the number of currently acquired mass spectrometry data reaches a preset quantity threshold, and the mass spectrometry data contains at least one standard sample data, all currently acquired mass spectrometry data are divided into sample data of the same batch; the preset quantity threshold can be flexibly set according to the actual detection process, for example, set to 30, 50 or 100, to ensure that the batch division will not result in too many batches due to the small number of samples, thus increasing the processing burden.

[0015] For example, when the preset quantity threshold is 30, if blank sample 1 to blank sample 10, standard sample 1, and test sample 1 to test sample 30 are injected sequentially during continuous detection, when test sample 19 is detected, the mass spectrometry data already contains one standard sample data and the preset quantity threshold is reached, so blank sample 1 to blank sample 10, standard sample 1, and test sample 1 to test sample 19 can be classified into the same batch. When blank samples 1 to 10, test samples 1 to 30, and standard sample 1 are sequentially injected during continuous detection, the mass spectrometry data will only contain one standard sample data when standard sample 1 is detected. At this point, the preset quantity threshold is reached, and blank samples 1 to 10, test samples 1 to 30, and standard sample 1 can be classified into the same batch. It should be noted that although the standard sample is located at the end of the batch in this example, and the judgment parameters cannot be determined before the standard sample detection is completed, the batch division is still valid. After the standard sample detection is completed, the test samples in the batch can be retrospectively judged based on the standard sample.

[0016] Preferably, the method proposed in this embodiment is pre-configured with a corresponding parameter lookup table, so that in step S100, the corresponding parameters can be called from the lookup table according to the different detection objects; in addition, the parameters can also be called from local configuration files, databases or method files. Using parameter lookup tables or configuration files can make parameter management more centralized and standardized, and facilitate maintenance and adjustment.

[0017] In this embodiment, different detection methods can include the same compound. That is, the same compound can be measured by different methods, while no duplicate compounds will appear within the same detection method. When both the detection method and the target compound are determined, the corresponding parameter information can be determined accordingly.

[0018] For example, the parameter reference table can be in the form of Tables 1-4, each table corresponds to a detection method. The daughter ion field can contain one or more characteristic ions. The number of daughter ions can be different for different compounds. The from_peak field is true or false, which is used to indicate whether the peak characteristics of the compound need special processing. Furthermore, the method proposed in this embodiment can also provide a graphical user interface (GUI), through which users can add, delete, and modify information in the parameter reference table. Specifically, this may include, but is not limited to, adding compound entries, modifying the mass-to-charge ratio of characteristic ions, adjusting the daughter ion list, and changing the from_peak identifier, thereby facilitating flexible management of parameter configurations for different detection methods and different target compounds.

[0019] Table 1 Phthalate Test Table

[0020] Table 2 Environmental Halogenated Hydrocarbons (env_test)

[0021] Table 3 Phthalate PHTH-8 MIX Table

[0022] Table 4 mzXMLtest table

[0023] S200: Based on characteristic ion information, extract ion flow from mass spectrometry data and locate candidate signal regions within a preset retention time range.

[0024] It should be noted that step S200 can extract the total ion current and / or extracted ion current from the mass spectrometry data based on the characteristic ions corresponding to the target compound. The preset retention time range is the allowable error range or time window. Step S200 can locate the candidate target peak or candidate signal region within this range, making full use of the characteristics of the target substance appearing near a specific retention time and having a characteristic mass-to-charge ratio signal, and separating the target-related signal from the complex background. In other alternative implementations, step S200 can be based on the extraction of a single feature ion, the combined extraction of multiple qualitative / quantitative ions, or the location of candidate signal regions can be achieved using an absolute time window or a relative time offset window.

[0025] S300: Based on the judgment condition information, the candidate signal region is judged, and when the judgment result is unclear or the judgment result indicates that there is a risk of interference, the peak shape characteristics of the candidate signal region corresponding to the mass spectrometry data are analyzed to identify abnormal signals.

[0026] In this embodiment, the determination of candidate signal regions includes a regularized determination of candidate signal regions based on one or more of the following determination conditions: retention time matching, peak area, peak height threshold, ion response relationship, existence of interference peaks and validity of integration results. It should be noted that the rule-based determination in step S300 can use dedicated rules established for the actual determination needs of the detection industry, rather than directly using the default integral and default peak recognition results of general software, so that the determination conditions can fit the method standards and determination criteria of specific detection projects. In other optional implementations, step S300 can use peak height determination, peak area determination, peak height and peak area combination determination, or single threshold, double threshold and multi-condition joint determination to determine the candidate signal region, so as to realize the application of different determination methods for different scenarios.

[0027] Please refer to Figures 2(a)-2(d). The peak shape characteristics of the candidate signal region include one or more of the following: peak width, peak shape symmetry, continuity, signal fluctuation, local extreme value distribution, and adjacent background relationship. The identified abnormal signals include one or more of the following: abnormal integral, pseudo peak, overlapping peak, irregular peak, and weak peak. As the main reason for misjudgment or omission by general software, abnormal signals can be effectively identified in step S300 through peak shape feature analysis.

[0028] Preferably, the method for analyzing the peak shape characteristics of candidate signal regions corresponding to mass spectrometry data can include one or more combinations of the following three methods: The first approach includes analyzing the peak shape characteristics of candidate signal regions based on preset mathematical rules. For example, the peak start and end points are determined by calculating the first derivative of the peak shape, the peak position and the existence of the shoulder peak are determined by the second derivative, and the degree of peak symmetry is quantified by calculating the symmetry factor. The second approach includes analyzing the peak shape characteristics of candidate signal regions based on statistical features, derivative features, and / or local window features. For example, calculating the mean, variance, skewness, kurtosis, and other statistical measures of the peak shape on the time axis, or calculating the local extreme value density, local signal-to-noise ratio, and other features within the sliding window. The third approach involves using a pre-trained classification model. The peak shape features of the candidate signal region are taken as input, and the analysis results of the peak shape features of the candidate signal region are output through the classification model. The classification model can be trained using labeled peak shape data as training data. The classification model can assist in the identification and classification of target peaks, interference peaks, abnormal peaks, or weak peaks, thereby improving the identification accuracy in complex situations. For peak shape features that are difficult to quantify or have ambiguous boundaries according to traditional rules, the machine learning model can learn the implicit patterns in a large number of labeled samples and give more stable and consistent judgment results than human experience.

[0029] S400: Based on the comprehensive judgment results and the identified abnormal signals, output the processing conclusions for the mass spectrometry data, and save the intermediate data of the mass spectrometry data during the detection process.

[0030] Specifically, the processing conclusion may include the conclusion of the existence of the target compound, and the specific content may include, but is not limited to, one or more of the following: the target compound exists, the target compound does not exist, the result is questionable, or manual verification is required. For example, when the rule judgment result meets all the judgment conditions and the peak shape characteristics are normal, step S400 can output the processing conclusion "the target compound exists"; when the rule judgment does not meet the conditions or the peak shape characteristics are identified as a false peak or interference peak, step S400 can output the processing conclusion "the target compound does not exist"; when the judgment result is unclear or there is a risk of interference but it cannot be confirmed, step S400 can output the processing conclusion "the result is suspicious" and / or "manual verification is required" accordingly. Thus, the method proposed in this embodiment can realize the automatic judgment of most normal samples, and can also leave the boundary cases to manual verification, thus taking into account both automation efficiency and judgment safety.

[0031] Intermediate data can include intermediate processing data, ion flow spectra, peak identification markers, integration start and end point information, judgment basis logs, and / or structured result files. Saving intermediate data facilitates subsequent traceability and auditing, meets the compliance requirements of the testing industry, and when the output conclusion is "result suspicious" or "requires manual review", the reviewer can quickly locate the problem based on this intermediate data, improving review efficiency.

[0032] It should be noted that the method proposed in this embodiment also has a differentiated processing mechanism for different sample types in the same batch. The differentiated processing mechanism does not require the samples to be injected in a fixed order, and can adapt to the flexible and varied sample sequence arrangement in actual testing. Specifically, when no standard sample data is detected in the sample data of the current batch, when detecting mass spectrometry data belonging to the current batch, the detection method is determined based on the mass spectrometry data according to the preset detection logic, and the parameter information is called based on the mass spectrometry data and / or the detection method. Finally, based on the judgment result of the candidate signal region and the analysis and identification result of the peak shape characteristics, each mass spectrometry data in the batch is determined as blank sample data, standard sample data or sample data to be tested.

[0033] In this embodiment, when blank sample data is detected, the determination and identification steps for blank sample data can be skipped. That is, the blank sample data is not used as the test sample to output the processing conclusion, and / or only background contamination, interference signal or system residue checks are performed on the blank sample data. It should be noted that the blank sample itself does not contain the target object, so there is no need to perform qualitative and quantitative determination of the target object. Skipping the processing steps can effectively save computing resources. When standard sample data is detected, characteristic ion information is determined based on the ion signal information and ion response intensity distribution information detected from the standard sample data. Furthermore, based on the retention time information, ion response information, peak shape characteristic information, signal-to-noise characteristic information, background characteristic information, inter-ion correlation characteristic information and / or model input features obtained by feature extraction from each of the standard sample data, judgment condition information for rule judgment and / or model judgment is determined. Once standard sample data has been detected in the current batch of sample data, when detecting mass spectrometry data belonging to the current batch, parameter information can be determined based on the mass spectrometry response characteristics of the standard sample data. Then, based on the parameter information, the data of each test sample in the same batch can be judged and identified. By combining the judgment results and the identified abnormal signals, a processing conclusion for each test sample data can be output. In this way, the test sample data can be used as a reference benchmark with the standard sample data in the same batch, so as to perform relative quantitative and qualitative judgment, offset the differences in instrument response between different batches, and ensure the consistency and comparability of the detection results.

[0034] As a verification of the method proposed in this embodiment, approximately 1800 historical and newly added mass spectrometry data were selected using this method, and automatic initial screening was performed based on rules such as retention time and abundance ratio. The processing results were compared with the results of manual review. The comparison results show that the method can filter approximately 91% of the data with a false negative rate of <0.2%, which is better than the false negative rate of manual review. This verification result fully demonstrates the reliability and safety of this method in automatic screening scenarios. For the more than 100 suspicious samples screened out by the rule layer, further identification was carried out by combining peak features and classification models, and compared with the results of manual judgment. The comparison results showed that about 20% of the suspicious samples could be further filtered, which fully demonstrated that the classification model has a certain ability to identify complex normal samples and can effectively reduce unnecessary manual review workload. The method proposed in this invention was deployed on a terminal device used for daily review. The terminal device was used to automatically read, analyze, and output batch mass spectrometry data. The software's running speed and stability were tested. The test results showed that the algorithm software could run smoothly without manual intervention. The processing time for a single data point was less than 1 minute, and the processing efficiency could reach about 30s / data point to 40s / data point. This met the timeliness requirements for batch data processing in daily testing work, and balanced the accuracy and efficiency of the processing.

[0035] Example 2 This embodiment provides a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the steps of the automatic processing method for compatible multi-source mass spectrometry data proposed in Embodiment 1 above.

[0036] It should be noted that computer-readable media, including both permanent and non-permanent, removable and non-removable media, can store information using any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage, quantum memory, graphene-based storage media or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0037] In summary, this invention proposes an automated processing method compatible with multi-source mass spectrometry data. The proposed solution is optimized for specific detection scenarios, avoiding problems such as inaccurate integration and peak identification deviation caused by default processing in general software. It significantly improves the accuracy, consistency, and automation level of mass spectrometry data processing, reduces the workload of manual review and reliance on experience judgment, and meets the practical application needs of the detection industry for high efficiency, stability, and automation.

[0038] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An automatic processing method compatible with multi-source mass spectrometry data, characterized in that, The method includes: The mass spectrometry data to be processed are acquired sequentially during the continuous detection process, and based on preset rules, multiple mass spectrometry data to be processed are divided into sample data of the same batch. Based on the sample name, sample sequence information, method report, file path, metadata fields, preset keywords, and / or ion response characteristics corresponding to the mass spectrometry data, the mass spectrometry data in the same batch are identified as blank sample data, standard sample data, or sample data to be tested; the sample data in the same batch contains at least one standard sample data. The detection method is determined based on the mass spectrometry data, and parameter information is invoked based on the mass spectrometry data and / or the detection method; the parameter information is determined based on the preset detection logic and / or the mass spectrometry response characteristics of standard sample data belonging to the same batch as the mass spectrometry data, and the parameter information includes at least characteristic ion information and judgment condition information; Based on the characteristic ion information, ion flow is extracted from the mass spectrometry data, and candidate signal regions are located within a preset retention time range; The candidate signal region is determined based on the determination condition information. When the determination result is unclear or the determination result indicates that there is a risk of interference, the peak shape characteristics of the candidate signal region corresponding to the mass spectrometry data are analyzed to identify abnormal signals. Based on the combined judgment results and the identified abnormal signals, a processing conclusion for the mass spectrometry data is output, and intermediate data during the detection process is saved.

2. The method according to claim 1, characterized in that, The method further includes: When no standard sample data is detected in the current batch of sample data, the detection method is determined based on the mass spectrometry data according to the preset detection logic, and the parameter information is called based on the mass spectrometry data and / or the detection method. After standard sample data has been detected in the current batch of sample data, parameter information is determined based on the mass spectrometry response characteristics of the standard sample data. Based on the parameter information, each of the test sample data in the same batch is judged and identified. By combining the judgment results and the identified abnormal signals, a processing conclusion for each of the test sample data is output.

3. The method according to claim 1, characterized in that, The method further includes: If blank sample data is detected, the blank sample data will not be used as the output processing conclusion for the sample to be tested, and / or, only the blank sample data will be checked for background contamination, interference signals or system residues. If standard sample data is detected, characteristic ion information is determined based on the ion signal information and ion response intensity distribution information detected from the standard sample data; and, based on the retention time information, ion response information, peak shape characteristic information, signal-to-noise characteristic information, background characteristic information, inter-ion correlation characteristic information and / or model input features obtained by feature extraction from each of the standard sample data, judgment condition information for rule judgment and / or model judgment is determined.

4. The method according to claim 1, characterized in that, The preset rules include one or more of the following: sample sequence continuity, data generation time, file directory, detection method identifier, standard sample location, sample name keywords, and / or quantity threshold.

5. The method according to any one of claims 1-4, characterized in that, The determination of the candidate signal region includes a regularized determination of the candidate signal region based on one or more of the following criteria: retention time matching, peak area, peak height threshold, ion response relationship, presence of interference peaks, and validity of integration results.

6. The method according to claim 1, characterized in that, The peak shape characteristics of the candidate signal region include one or more of the following: peak width, peak shape symmetry, continuity, signal fluctuation, local extremum distribution, and relationship with adjacent background. The identified anomalous signals include one or more of the following: anomalous integrals, spurious peaks, overlapping peaks, irregular peaks, and weak peaks.

7. The method according to claim 1 or 6, characterized in that, The method for analyzing the peak shape characteristics of the candidate signal region corresponding to the mass spectrometry data includes: The peak shape characteristics of the candidate signal region are analyzed based on preset mathematical rules; The peak shape characteristics of the candidate signal region are analyzed based on statistical characteristics, derivative characteristics, and / or local window characteristics. And / or, using a pre-trained classification model, the peak shape features of the candidate signal region are taken as input, and the analysis results of the peak shape features of the candidate signal region are output through the classification model.

8. The method according to claim 1, characterized in that, The processing conclusion includes one or more of the following: the target compound exists, the target compound does not exist, the result is questionable, or manual verification is required. The intermediate data includes intermediate processing data, ion flow spectrum, peak identification markers, integration start and end point information, judgment basis logs, and / or structured result files.

9. A computer-readable storage medium, characterized in that, It stores executable instructions for use by a processor to implement the automatic processing method for compatible multi-source mass spectrometry data as described in any one of claims 1-8.

Images