Method for realizing accurate identification of chemical components in plant complex matrix based on in-source pyrolysis multi-fragment information

Abstract

The invention provides a method for realizing accurate identification of chemical components in a plant complex matrix based on in-source pyrolysis multi-fragment information, is explained by taking identification of pesticides in the plant matrix as an example, and belongs to the technical field of accurate identification of chemical components. Comprising the following steps: S1, preparing a to-be-detected sample solution; s2, collecting sample data; s3, extracting an ion chromatogram, and carrying out peak detection and annotation; s4, identifying fragment ions in the source; and S5, identifying the compound based on the in-source fragment ions. The method can be used for accurately identifying the compounds in the plant complex matrix, matching factors of the compounds can be improved by identifying fragment ions in a source, false positive results can be reduced, and the performance of a developed strategy is comprehensively researched by using a complex data set; the problems of false positive and inaccurate identification in a traditional molecular ion identification strategy can be effectively solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of accurate identification of chemical components, and specifically relates to a method for realizing accurate identification of chemical components in complex plant matrices based on information of multiple fragments of intra-source pyrolysis. Background Art

[0002] The accurate identification of compounds in complex plant matrices is crucial to understanding plant metabolism, developing natural medicines, ensuring food safety and conducting environmental monitoring. It not only promotes basic scientific research, but also provides important support for applications in agriculture, medicine and environmental protection. However, there are many challenges in actual analysis, including the complex composition of plant matrices, the presence of multiple compounds, the low concentration of target compounds and the influence of interfering substances. These factors increase the difficulty of identification and may lead to misjudgment or missed detection. Therefore, corresponding technologies and methods are urgently needed to improve the accuracy and reliability of chemical component identification in complex plant matrices. Traditional compound identification methods mainly include targeted analysis and non-targeted analysis. Targeted analysis focuses on the detection of known compounds, usually with high sensitivity and accuracy, and is usually achieved through tandem mass spectrometry.

[0003] The development of ultra-high performance liquid chromatography-high resolution mass spectrometry (UHPLC-HRMS) has promoted the application of non-targeted analytical methods, such as metabolomics and screening of illegal additives. Compared with targeted analytical methods, a significant advantage of non-targeted analytical technology is that it does not rely on preset target substances, thus providing comprehensive and unbiased results for potential compounds. Thanks to the separation capabilities of the chromatographic and mass spectrometric dimensions, UHPLC-HRMS can separate and characterize tens of thousands of compounds, making it possible to use MS / MS information for high-throughput identification of compounds.

[0004] Since the composition of plant matrices is extremely complex and contains a variety of compounds, UHPLC-HRMS can provide a large amount of data on related compounds in non-targeted analysis. Effective data processing methods are essential for extracting information about compounds in samples. Currently, there are several commercially available analysis software for processing data obtained from UHPLC-HRMS, such as Progenesis QI, MarkerView, and Compound Discoverer. In addition, some freely available tools have been developed for UHPLC-HRMS data analysis, such as XCMS. The extensive development of these software has provided great help for compound extraction and identification in non-targeted metabolomics. Researchers combine MS / MS mass spectra and their precursor ions for compound identification. However, the existence of false positive identifications requires manual verification, which undoubtedly consumes time in routine analytical work.

[0005] Compounds can often generate multiple in-source fragment ions during mass spectrometry analysis, and these in-source fragment ions can also be used to obtain MS / MS spectra during data acquisition. In practice, researchers may mistakenly label these in-source fragment ions as independent precursor ions, leading to unacceptable compound identification results, a problem that cannot be solved by high-precision m / z values. Accurately identifying the in-source fragment ions of a compound can not only reduce the number of candidate compounds, but also improve the quality of compound identification. Obviously, the more in-source fragment ions identified from a compound, the more likely it is that the compound is present in the analyzed sample. However, the identification of in-source fragment ions is still insufficient in current UHPLC-HRMS data analysis tools, which means that the auxiliary role of in-source fragment ions in compound identification has not been fully appreciated. Summary of the invention

[0006] In view of this, the purpose of the present invention is to provide a method for accurately identifying chemical components in complex plant matrices based on in-source cleavage multi-fragment information. The method can be used to accurately identify compounds in complex plant samples, and is illustrated using the identification of pesticides in plant matrices as an example. The identification of in-source fragment ions can not only improve the matching factor of the compound, but also reduce false positive results. The performance of the developed strategy is comprehensively studied using complex data sets, which can effectively solve the problems of false positives and inaccurate identification in traditional molecular ion-based identification strategies.

[0007] In order to achieve the above object, the present invention provides the following technical solutions:

[0008] The present invention provides a method for realizing accurate identification of chemical components in complex plant matrices based on in-source pyrolysis multi-fragment information, comprising the following steps:

[0009] S1. Prepare the sample solution to be tested;

[0010] S2. Collect sample data;

[0011] S3. Extract ion chromatograms for peak detection and annotation;

[0012] S4. Identify the fragment ions in the source;

[0013] S5. Identify compounds based on in-source fragment ions.

[0014] Furthermore, the specific method of S1 is:

[0015] The sample to be tested is subjected to ultrasonic extraction using a mixture of methanol and water, and then centrifuged to obtain a supernatant, and a mixed pesticide solution is added to the supernatant.

[0016] Furthermore, the specific method of S2 collecting sample data is:

[0017] The sample solution was subjected to UHPLC-HRMS analysis, and the compounds were separated using a Waters Acquity UPLC HSS T3 column, 50 mm × 2.1 mm, with a particle size of 1.8 μm;

[0018] The column temperature was set at 35°C;

[0019] The mobile phase consisted of two parts: phase A was 0.1% formic acid aqueous solution and organic phase, and phase B was 0.1% formic acid acetonitrile solution; the flow rate was 0.2 mL / min;

[0020] The elution gradient was set as follows: 0 min, 97% A; 1 min, 97% A; 1.5 min, 85% A; 23 min, 50% A; 28 min, 30% A; 33 min, 2% A; 37 min, 2% A; 37 min, 97% A; 40 min, 97% A;

[0021] In terms of mass spectrometry parameters, the psi value of GS1 and GS2 was set to 50, and the curtain gas pressure was 30 psi;

[0022] The ion source temperature was set to 500°C;

[0023] The values ​​of ISVF and DP were set to 5500 V and 80 V, respectively;

[0024] Using DDA data acquisition mode, MS 1 The scanning range is 100-1000Da, and the resolution is 15000;

[0025] The scan range of MS / MS was set to 50–1000 Da;

[0026] The mass tolerance of the precursor ion was 100 mDa;

[0027] The collision energy of the product ions was 35 V, and the collision assist voltage was 15 V;

[0028] The dynamic exclusion time was set to 5 seconds.

[0029] Furthermore, the S3 is performed in the AntDAS-DDA platform.

[0030] EIC peak detection includes EIC construction, peak detection and ion annotation. The EIC of a compound is constructed based on its high-precision m / z value. The baseline drift from the background ions is corrected by a baseline drift correction algorithm based on the local minimum. A multi-scale smoothing strategy is used to extract compound features, and features with low signal-to-noise ratios are removed. The detected peaks will be characterized by retention time, high-precision m / z values, and peak area. Peak annotation uses high-precision m / z values ​​to identify adduct ions (such as +Na, +K,

[0031] +NH4), isotopic ions (such as +1, +2, +3), and neutral losses (such as -H2O, -2H2O) are implemented in the AntDAS-DDA software and can be obtained through http: / / www.pmdb.org.cn / antdasdda.

[0032] Furthermore, the method for S4 to identify the fragment ions in the source is:

[0033] i) selecting an EIC peak m having an MS / MS spectrum and extracting ions from the MS / MS spectrum to search for potential in-source fragment ions;

[0034] ii) if the high-precision m / z value of another EIC peak n matches an ion in the MS / MS spectrum in step i), then use this peak for in-source fragmentation checking;

[0035] iii) The Pearson correlation coefficient is used to characterize the similarity of peak values ​​m and n, which is calculated from their elution curves; the calculation formula is:

[0036]

[0037] Among them, S m , n is the Pearson correlation coefficient between EIC peaks m and n; c m and c n are the chromatogram vectors corresponding to EIC peaks m and n, respectively; and is the average of cm and cn; the superscript T indicates the transpose of the vector; ions with a Pearson correlation coefficient exceeding 0.975 will be marked as in-source fragment ions from the same compound;

[0038] iv) Construct MS for each compound based on the identified in-source fragment ions 1 Spectra, ultimately the MS of the in-source fragment ions constructed 1 Spectra and MS / MS spectra were used for compound identification.

[0039] Furthermore, the method for identifying compounds based on the fragment ions in the source in S5 is:

[0040] i) introducing the precursor ions into a pesticide database to find candidates that meet the m / z tolerance;

[0041] ii) Compare the in-source fragment ions of the unknown compound with the fragment ions of the candidate in the library. Specifically, compare the MS / MS spectra of the in-source fragment ions with the corresponding spectra in the library to calculate the similarity or match factor MF, as shown below:

[0042]

[0043] Among them, MF i is the i-th in-source fragment ion of the unknown compound; p i is the normalized MS / MS spectrum corresponding to the ion; r j is the normalized MS / MS spectrum of the jth in-source fragment ion of the reference compound in the library; the precursor ion of pi should be the same as r j The precursor ions are the same;

[0044] MF i Values ​​greater than 0.5 were considered acceptable matches and used in subsequent analyses;

[0045] iii) If in MS 1 If there are multiple in-source fragment ions in the spectrum, the MS 1 The matching factor MF of the spectrum is marked as MF MS1 , used for compound identification;

[0046] iv) Construct a total match factor TMF for each candidate compound, calculated according to the following formula:

[0047] TMF=MFMS1+ΣMF

[0048] Among them, MF MS1 It is MS 1 Matching factor of the spectrum, MF i is the matching factor of the i-th in-source fragment ion, and the candidate compound with the highest TMF will be selected as the identification result;

[0049] In practical applications, the low content of a compound may reduce the number of in-source fragment ions that can be generated in the high-resolution mass spectrometry DDA mode, and at least two in-source fragment ions should be matched to ensure effective compound identification.

[0050] Preferably, the method for constructing the pesticide database is:

[0051] The raw data obtained by UHPLC-HRMS analysis of the compound standards were imported into AntDAS for data analysis, including EIC construction, chromatographic peak extraction, chromatographic peak annotation, fragment ion identification, MS 1 and MS / MS spectrum construction to obtain the MS of the compound standards 1 and MS / MS spectrum information;

[0052] After manual confirmation, the Chinese and English names, CAS numbers, molecular masses, SMILES numbers, molecular formulas, precursor ions, collision voltages, instrument information, and mass spectrum information of the compounds were recorded and integrated to complete the construction of the database.

[0053] Contains at least the following beneficial technical effects:

[0054] The method of the present invention can be used to accurately identify compounds in complex plant samples, especially pesticides in plant matrices. The identification of in-source fragment ions can not only improve the matching factor of the compound, but also reduce false positive results. The performance of the developed strategy is comprehensively studied using complex data sets, which can effectively solve the false positive and inaccurate identification problems in traditional molecular ion-based identification strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 Workflow diagram for the development strategy: (a1) total ion chromatogram (TIC) of samples in dataset 1; (a2) extracted ion chromatogram (EIC) of m / z 309.0189Da; (a3) ​​corresponding EIC of each ion; (b1) identification of in-source fragment ions; (b2) network relationship diagram of in-source fragment ions and their MS / MS spectra; (c1)-(c4) compound identification based on in-source fragment ions; (c5) identified compounds.

[0056] Figure 2 Illustrations of compound screening and identification in Dataset 1: (a) Number of compounds identified at different addition levels in Dataset 1; (b) Distribution of TMF at each addition level; (c) TMF of carbofuran compounds at different addition levels; (d) MS of carbofuran compounds at an addition level of 300 μg / g 1 Spectrum; (e) In-source fragment ion network of carbofuran compounds; (f) Chromatographic elution curve of carbofuran compound ions; (g) MS 1 Detailed information of all fragment ions in the spectrum; (h) and (i) are based on the constructed MS of m / z 165.0903Da. 1 and MS / MS spectrum matching results.

[0057] Figure 3Illustration of the identification of real sample compounds in Dataset 2: (a) MS of methoxyfenozide constructed 1 Spectrum; (b) network of its in-source fragment ions; (c) chromatographic elution curve of in-source fragment ions; (d) detailed information of the matched in-source fragment ions; (eg) matching ions and MS of m / z 313.1540 and m / z 391.1976 1 Schematic representation of the spectrum; (f) Confirmation with standard compounds.

[0058] Figure 4 Statistical analysis of the screened compounds: (a) the number of samples collected and the number of samples that screened out at least one pesticide; (b) sunburst analysis of pesticides in different plant matrices.

[0059] Figure 5 Comparison diagram with the classical strategy: (a) classical compound identification based on molecular ion MS / MS spectra; (b) compound identification based on the developed in-source fragment ion recognition strategy; (c) inaccurate identification in the classical method; (d) identification results of the developed strategy.

[0060] Figure 6 Figure 2 shows the effect of parameters on the performance of the development strategy, with low addition level (a), medium addition level (b) and high addition level (c).

[0061] Figure 7 This is a characteristic detection result diagram of the compound iprodione at the lowest additive level (30 μg / g).

[0062] Figure 8 Graph comparing the false positive rate (a) and matching factor (b) between the detection method of the present invention and the traditional strategy.

[0063] Fig. 9 This is the statistical distribution diagram of fragment ions in 478 pesticide sources. DETAILED DESCRIPTION

[0064] The technical solution of the present invention is further described below through the accompanying drawings and embodiments.

[0065] Unless otherwise defined, technical or scientific terms used in the present invention shall have the common meanings understood by one having ordinary skills in the field to which the present invention belongs.

[0066] It is obvious to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the main idea or essential features of the present invention. Therefore, from all points of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention, and any reference numerals in the claims should not be regarded as limiting the claims involved.

[0067] In addition, it should be understood that although this specification is described according to the implementation modes, not every implementation mode includes only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation modes that can be understood by those skilled in the art. These other implementation modes are also covered within the protection scope of the present invention.

[0068] It should also be understood that the specific embodiments described above are only used to explain the present invention, and the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, can make equivalent replacements or changes based on the technical solutions and inventive concepts of the present invention, which should be covered by the protection scope of the present invention / invention.

[0069] Example 1

[0070] 1. Targeted Dataset Verification

[0071] Eighteen pesticide standards were added to chrysanthemum tea matrix at different spiking levels. The detailed information of the 18 pesticides is shown in Table 1.

[0072] Table 1 Detailed information of 18 pesticides

[0073]

[0074] 1.1 Preparation of pesticide mixture

[0075] The stock solutions of each pesticide in Table 1 were prepared at a concentration of 8 μg / mL. Subsequently, a series of mixed solutions were prepared by mixing these stock solutions in different amounts. In each mixed solution, the concentrations of the 18 pesticides were kept consistent. The concentrations of the pesticides in the mixed solutions ranged from 40 ng / mL to 400 ng / mL in increments of 40 ng / mL. Finally, a total of 10 mixed solutions were prepared.

[0076] 1.2 Preparation of chrysanthemum tea sample solution

[0077] 20 mg of chrysanthemum tea was weighed into a 2 mL Eppendorf tube, and then 1.5 mL of extraction solvent (methanol / water = 70 / 30; v / v) was added. After 2 minutes of vortex mixing, ultrasonic treatment was performed at room temperature for 30 minutes. Subsequently, the Eppendorf tube was centrifuged at 13000 rpm for 10 minutes. Next, 0.5 mL of supernatant was transferred to another Eppendorf tube, and 0.5 mL of pesticide mixed solution was added for UHPLC-HRMS analysis, and the relative concentration range of the pesticide added solution was controlled to be 30 μg / g to 300 μg / g, with an increment of 30 μg / g. Finally, 10 concentrations of pesticide added solutions were constructed to evaluate the performance of the developed strategy.

[0078] 1.3 Sample data collection

[0079] The parameters are as follows:

[0080] Each pesticide-added solution was analyzed on an AB SCIEX instrument to collect data. Compounds were separated using a WatersAcquity UPLC HSS T3 column (50 mm × 2.1 mm, particle size 1.8 μm). The column temperature was set to 35 ° C. The mobile phase consisted of two parts: an aqueous phase (phase A, 0.1% formic acid in water) and an organic phase (phase B, 0.1% formic acid in acetonitrile) with a flow rate of 0.2 mL / min. The elution gradient was set to: 0 min, 97% A; 1 min, 97% A; 1.5 min, 85% A; 23 min, 50% A; 28 min, 30% A; 33 min, 2% A; 37 min, 2% A; 37 min, 97% A; 40 min, 97% A.

[0081] In terms of mass spectrometry parameters, the psi values ​​of GS1 and GS2 were set to 50, and the curtain gas pressure was 30 psi. The ion source temperature was set to 500 °C. The values ​​of ISVF and DP were set to 5500 V and 80 V, respectively. The DDA data acquisition mode was used, and the MS 1 The scan range was 100-1000Da with a resolution of 15000. The scan range of MS / MS was set to 50-1000Da. The mass tolerance of the precursor ion was 100mDa. The collision energy (CE) of the product ion was 35V and the collision assist voltage (CES) was 15V. The dynamic exclusion time was set to 5 seconds.

[0082] 1.4 Extracting ion chromatograms for peak detection and annotation

[0083] EIC peak detection includes EIC construction, peak detection and ion annotation. The EIC of a compound is constructed based on its high-precision m / z value. The baseline drift from the background ions is corrected by a baseline drift correction algorithm based on the local minimum. Compound features are extracted using a multi-scale smoothing strategy, and features with low signal-to-noise ratios are removed. The detected peaks will be characterized by retention time, high-precision m / z values, and peak area. Peak annotation uses high-precision m / z values ​​to identify adduct ions (such as +Na, +K, +NH4), isotope ions (such as +1, +2, +3), and neutral losses (such as -H2O, -2H2O). These functions are implemented in our AntDAS-DDA software and can be obtained through http: / / www.pmdb.org.cn / antdasdda.

[0084] Examples of peak detection and annotation are shown in Figure 1 As shown in a1-a3. The raw data file from high-resolution mass spectrometry (HRMS) (Figure a1) is used to automatically perform EIC construction and peak detection (Figure a2). As shown in the figure, two peaks were detected in the EIC of m / z 309.0189. The peak annotation results were based on high-precision m / z values. m / z 309.0190, 310.0221 and 311.0152 in (Figure a3) were annotated as isotope ions [M+H]+, [M+1+H]+ and [M+2+H]+, respectively.

[0085] 1.5 Identification of in-source fragment ions

[0086] The identification of in-source fragment ions was performed based on the data-dependent acquisition mode of UHPLC-HRMS. The fragment ions of the compounds were identified using the MS / MS spectra by the following steps:

[0087] i) For an EIC peak m with an MS / MS spectrum, ions are extracted from the MS / MS spectrum to search for potential in-source fragment ions.

[0088] ii) If the high-precision m / z value of another EIC peak n matches an ion in the MS / MS spectrum in step i), then this peak is used for in-source fragmentation checking.

[0089] iii) The Pearson correlation coefficient is used to characterize the similarity of peaks m and n, which is calculated from their elution curves:

[0090]

[0091] Among them, s m,n is the Pearson correlation coefficient of EIC peaks m and n, c m and c nare the chromatogram vectors corresponding to EIC peaks m and n, respectively, and It is c m and c n The superscript T indicates the transpose of the vector. Ions with a Pearson correlation coefficient exceeding 0.975 are labeled as in-source fragment ions from the same compound.

[0092] iv) Construct MS for each compound based on the identified in-source fragment ions 1 The MS of the in-source fragment ions will be constructed. 1 Spectra and MS / MS spectra were used for compound identification.

[0093] In-source fragment ion identification strategy Figure 1 As shown in b1-b2, Figure b1 shows some EIC peaks detected in the analyzed samples, and the last column marks the EIC peaks with MS / MS spectra, which are used to find in-source fragment ions. Taking the EIC peak of m / z 309.0190 as an example, its MS / MS spectrum is shown in Figure b2. In addition to the molecular ion (m / z 309.0190) and the summed ion +Na (m / z 331.0011), an ion of m / z 156.0203 was also found. The Pearson correlation coefficient between m / z 156.0203 and the EIC peak of m / z 309.0190 is 0.996, which meets the threshold of 0.975. Therefore, the ion m / z 156.0203 is marked as an in-source fragment ion. Finally, the MS of the compound was constructed. 1 The MS / MS spectrum of the in-source fragment ions was obtained simultaneously.

[0094] 1.6 Identification of compounds based on in-source fragment ions

[0095] The in-source fragment ions of a compound directly reflect the dissociation of its chemical structure during the ionization process. Although these fragment ions can theoretically be used for pesticide identification, few data analysis tools take all in-source fragment ions into account. In our method, compound MS 1 The MS / MS spectra of all ions in the spectrum are identified, including the following steps:

[0096] i) The precursor ions are introduced into a pesticide database to find candidates that meet the m / z tolerance (eg 0.01 Da).

[0097] ii) Compare the in-source fragment ions of the unknown compound with the fragment ions of the candidate in the library. Specifically, the MS / MS spectra of the in-source fragment ions are compared with the corresponding spectra in the library to calculate the similarity or match factor (MF) as follows:

[0098]

[0099] Among them, MF i is the i-th in-source fragment ion of the unknown compound. i is the normalized MS / MS spectrum corresponding to this ion. j is the normalized MS / MS spectrum of the jth in-source fragment ion of the reference compound in the library. It is worth noting that p i The precursor ion should be j The precursor ion of MF is the same. i Values ​​greater than 0.5 were considered acceptable matches and used in subsequent analyses.

[0100] iii) If in MS 1 If there are multiple in-source fragment ions in the spectrum, the MS 1 The match factor (MF) of the spectrum, and is labeled MF MS1 , used for compound identification.

[0101] iv) Construct a total match factor (TMF) for each candidate compound, calculated according to the following formula:

[0102] TMF=MFMS1+ΣMF

[0103] Among them, MF MS1 It is MS 1 Matching factor of the spectrum, MF i is the match factor of the i-th in-source fragment ion. The candidate compound with the highest TMF will be selected as the identification result. However, in practical applications, the low content of the compound may reduce the number of in-source fragment ions that can be generated in the high-resolution mass spectrometry DDA mode, requiring at least two in-source fragment ions to be matched to ensure effective compound identification.

[0104] Figure 1 Figures c1-c5 show the compound identification strategy based on in-source fragment ions. Figure c1 describes the MS 1 Spectrum. The ion at m / z 309.0190 is the molecular ion of the compound and is used to search for potential candidates in the database. MS 1 There are three in-source fragment ions in the spectrum that can obtain MS / MS spectra during data acquisition, including m / z 156.0203, 309.0190, and 313.0128. The MS / MS spectra of these three ions are shown in Figures c2-c4. The MS / MS spectra of these three ions are compared with the MS / MS spectra of each candidate ion to calculate the TMF. In Figure c5, the MS / MS spectra between the unknown compound and the candidate compound diflubenzuron are shown in Figure c2-c4. 1The comparison with MS / MS spectra is shown in Figures c1-c4. The matching factors (MF) of the fragment ions of m / z 156.0203, 309.0190 and 313.0128 are 0.991, 0.956 and 0.983, respectively. 1 The MF of the spectrum was 0.955, and the final TMF was 3.885.

[0105] 1.7 Test results

[0106] The performance of the developed method was first evaluated by identifying standard materials spiked into a complex plant matrix, chrysanthemum.

[0107] Figure 2 a Provides the number of standard compounds identifiable at each spike level.

[0108] Table 2 lists the detailed identification results of the 18 added compounds. At levels above 30 μg / g, all 18 added pesticide standards were successfully identified. At the lowest added concentration level (30 μg / g), only the pesticide standard of isoprodinil could not be identified because no MS / MS spectrum was obtained for this compound at the lowest added level during data collection (see Figure 7 ).

[0109] Figure 2 b summarizes the TMF distribution characteristics of the identified compounds at each addition level. The TMF values ​​of all identified compounds are much greater than 1. Taking the lowest addition level as an example, the median TMF of the identified compounds is 1.921, indicating that MS 1 Several in-source fragment ions in the spectrum match the reference ions. Figure 2 b It can also be seen that the TMF gradually increases with the increase of the addition level. One possible reason is that the number of in-source fragment ions for which MS / MS spectra can be obtained increases in DDA mode.

[0110] Figure 2 c A detailed description is given using the added pesticide standard carbofuran as an example. It can be clearly seen that as the level of additive increases, the TMF shows an overall upward trend. Multiple in-source fragment ions match the reference ions, making the TMF of all added levels exceed 2.5. Figure 2 d to Figure 2 i shows the identification process based on in-source fragment ions. Figure 2 d shows the constructed carbofuran MS at the 300 μg / g addition level. 1 Spectrum, in which the in-source fragment ions m / z 222.1118, 165.0903 and 123.0437 were identified using the developed strategy. The ion network of these ions is shown in Figure 2. Figure 2e. It can be found that during the peak detection and annotation process, four ions were labeled as [M+H]+ ions, which may be mistakenly identified as different compounds in traditional compound identification methods. However, our strategy can be used to identify different compounds through MS / MS spectra and EIC ( Figure 2 f) Identify in-source fragment ions. For example, ions of m / z 222.1118, 165.0903, and 123.0437 can be found in the MS / MS spectrum of ion m / z 267.1695. In addition, the EIC similarity between these ions is greater than 0.975, and ultimately in the MS 1 The ions in the spectrum were identified as in-source fragment ions.

[0111] Figure 2 g shows all ions identified after in-source fragmentation, of which 6 ions obtained MS / MS spectra. The MF of each ion is shown in Figure 2 g, ranging from 0.598 to 0.990. 1 The MS / MS spectra were compared with those in the library, and the results were as follows: Figure 2 Finally, the TMF value obtained is 6.239, which is shown in Figure 2 In the last column of g, the value is calculated based on all matching MS / MS spectra and MS 1 Finally, the carbofuran compounds were successfully confirmed with the standards. Figure 2 The results in show that the developed method can successfully identify compounds with the help of in-source fragment ions.

[0112] Classical compound identification was also performed for this compound based on its molecular ion m / z 267.1695, and the MF of its MS / MS spectrum was 0.983. Comparison between the development strategy (6.239) and the classical strategy (0.983) showed that the MF could be improved by 6 times.

[0113] Table 2 Overall matching factors of 18 standard compounds added to chrysanthemum matrix

[0114]

[0115]

[0116] Example 2

[0117] Conducting a real experiment

[0118] A total of 80 real samples were purchased from the local market in Yinchuan, derived from four plant matrices, including chrysanthemum, rose, green tea, and wolfberry bud tea. Twenty samples were collected from each plant matrix and ground into powder for metabolite extraction. The sample preparation method was the same as that of chrysanthemum tea in Example 1. At the same time, the parameters of UHPLC-HRMS in Dataset 1 were used for data collection; the detection method was the same as 1.4-1.6 in Example 1.

[0119] Subsequently, the performance of the developed strategy was studied using pesticide screening and identification as examples using real samples in Dataset 2. A database of more than 470 compounds was then constructed for pesticide screening. During the data collection process, the in-source fragment ions of each compound were collected and stored in the database, and the construction of the database can be done using existing methods.

[0120] Figure 3 The process of pesticide screening in complex plant samples was specifically demonstrated, taking the screening of methoxyfenozide in chrysanthemum plant matrix as an example. Figure 3 a provides a MS based on the developed strategy 1 In the spectrum, in-source fragment ions such as m / z 313.1540 and 149.0591 can be observed. In addition, adduct ions such as m / z 391.1976, which are annotated as [M+Na]+, can also be observed in the MS 1 Found in the spectrum. Figure 3 The ion network in b shows that m / z 313.1540 and 149.0591 are also fragment ions in the MS / MS spectrum of m / z 369.2164. Interestingly, m / z 313.1540 and 149.0591 can also be detected in the MS / MS spectrum of ion m / z 414.2736. Figure 3 c provides the EICs of all relevant ions, where it can be seen that the elution profiles of all ions are very similar.

[0121] By introducing the ion of m / z 369.2164 into the pesticide database, a candidate compound, methoxyfenozide, was identified. Figure 3 d provides MS 1 MF of each in-source fragment ion in the spectrum. It can be found that 5 ions match the reference ions in the database, with MF ranging from 0.718 to 0.999. In addition, the constructed MS 1 The spectrum obtained a MF of 0.980. Figure 3 e-3g shows MS 1 and MF of several in-source fragment ions. 1 Spectrum( Figure 3e) is almost identical to the reference spectrum in the library (0.980), indicating that the in-source fragment ions of the compound can be accurately identified. In-source fragment ions such as m / z 313.1540 and 391.1976 obtained 0.999 ( Figure 3 f) and 0.989( Figure 3 To further verify the accuracy of compound identification, standard compounds were analyzed under the same UHPLC-HRMS conditions. Figure 3 h. It can be clearly seen that the retention time of the standard is 26.69 min, which is very close to the retention time of 26.76 min in the analyzed sample. Figure 3 The results in show that accurate screening and identification of pesticides in complex plant matrices can be achieved using the developed strategy.

[0122] With the help of this developed strategy, nine pesticides were screened and identified from 80 samples, including carbendazim, acetamiprid, imidacloprid, methyl thiophanate, thiamethoxam, methoxyfenozide, fluopyram, carbofuran and thiacloprid. Figure 4 a provides details on the number of compounds in each plant matrix. Specifically, 7 pesticides were detected in chrysanthemum, followed by wolfberry bud tea (5 pesticides), rose tea (2 pesticides), and green tea (2 pesticides). Carbendazim was mainly found in rose samples, while acetamiprid, imidacloprid, methylthiophanate, and thiamethoxam were mainly found in wolfberry bud tea samples. As for methoxyfenozide, it was mainly found in chrysanthemum samples ( Figure 4 b) All the screened pesticides were confirmed by standards and detailed information on the specific pesticides in each sample is provided in Table 3.

[0123] Table 3 Detailed information of pesticides screened from 80 real samples

[0124]

[0125]

[0126]

[0127]

[0128]

[0129] Example 3

[0130] 1. Comparative Experiments with Traditional Methods

[0131] When using UHPLC-HRMS technology for complex matrix analysis of plants, compound identification based on high-precision m / z values ​​and corresponding MS / MS spectra often encounters two challenges. The first is the emergence of multiple candidates that all meet the mass-to-charge ratio tolerance of [M+H]+ and the MF of MS / MS, which may confuse analysts in selecting appropriate candidates. The second is that in-source fragment ions with MS / MS spectra may be mistakenly identified as different compounds. With the help of the developed strategy, these two problems can be reasonably solved.

[0132] Figure 5 A graphical comparison of the developed compound identification strategy with the traditional strategy is shown, where Figure 5 a and Figure 5 b provides examples of compound identification under the traditional strategy and the development strategy, respectively. Figure 5 a provided the MS / MS spectrum of the molecular ion m / z 329.0809. By comparing the MS / MS spectrum with the MS / MS spectrum of pyrifosamide in the database, a MF of 0.926 was obtained, indicating that the correct compound identification results can be obtained by traditional methods.

[0133] The developed strategy provides an MS 1 The spectrum contains three ions with MS / MS spectra, namely the molecular ion m / z 329.0809 and two in-source fragment ions m / z 302.0704 and 304.0671. Figure 5 The MS / MS spectra of these three ions are shown in b, and the constructed MS 1 In addition, the spectra of the in-source fragment ions in the database were investigated and all 3 MS / MS spectra were found. A comparison of the constructed spectra with those in the database can be found in Figure 5 b. The MFs of the in-source fragment ions of m / z 302.0704 and 304.0671 were 0.951 and 0.913, respectively. 1 The MF obtained for the spectrum was 0.996. Finally, the developed strategy provided a TMF of 3.786, which could greatly facilitate compound identification by providing a MF 4 times larger than the traditional strategy.

[0134] Another example of reducing false positive screening through the developed strategy is shown in Figure 5 on the right side. Figure 5c shows compounds screened by the traditional strategy using the molecular ion and its MS / MS spectrum. In this case, the compound imazethapyr was matched by providing a molecular ion m / z tolerance of 0.0086 Da and an MS / MS spectrum MF of 0.656, which might be accepted by the analyst without manual verification.

[0135] However, the developed strategy will use in-source fragment ions for compound identification. It was found that the imidazolin compound provided three in-source fragment ions and their MS / MS spectra in the library ( Figure 5 d), but these fragment ions could not be found in the analyzed samples ( Figure 5 d). Finally, the compound obtained a TMF of 0.67 by the developed strategy and was marked as an invalid identification. Standard confirmation demonstrated the success of the developed strategy, indicating that the developed strategy can avoid false positive screening.

[0136] Figure 5 The results in confirm the success of using in-source fragment ions for compound identification. Statistical analysis of MF for reducing false positive screening and improving compound identification was performed using Dataset 1. The results are shown in Figure 8 As shown. It can be found that the average false positive rate of the traditional screening strategy is above 134.8%. With the help of the developed strategy, the false positive screening results can be reduced to 15.8%, a reduction of 119.0% ( Figure 8 a) In addition, MF from development strategy and traditional strategy such as Figure 8 As shown in (b), it can be found that the average MF of the developed strategy is 4.14, which is 4.3 times that of the traditional method (0.96).

[0137] The performance of the developed strategy relies on in-source fragment ions. The more in-source fragment ions matched, the higher the confidence in compound identification. A study of 478 pesticides showed that more than 280 compounds were identified in MS 1 At least two fragment ions can be provided in the spectrum ( Fig. 9 ). For compounds that only provide a single MS / MS spectrum, the developed strategy will degenerate into traditional pesticide screening and identification.

[0138] 2. Influence of parameters

[0139] In practice, the TMF calculated by the developed strategy depends on two parameters, namely the MS / MS spectrum similarity threshold and the number of matched in-source fragment ions. These two parameters were studied in depth in the samples of Dataset 1. Three spiking concentration levels were studied: low spiking level 30 μg / g, medium spiking level 150 μg / g and high spiking level 300 μg / g. False positive and false negative results were analyzed and reported in Figure 6The relevant information is provided in . The different lines mark the similarity thresholds of the MS / MS spectra, while the numerical values ​​in each sub-graph are the thresholds of the number of matching in-source fragment ions. It should be noted that when the numerical value is 1, the developed strategy is equivalent to the traditional strategy.

[0140] By increasing the MS / MS spectrum similarity threshold, false positive results can be significantly reduced, especially when the number of matched in-source fragment ions is 1. For example, Figure 6 The results in a show that by increasing the similarity threshold of the MS / MS spectra from 0.6 to 0.8, the false positive rate can be reduced from 186.7% to 53.3%. However, as the additive level increases, the ability to reduce false positive results decreases. Taking the 300μg / g additive level as an example, when the threshold is set from 0.6 to 0.8, the false positive rate is reduced from 57.9% to 42.1%. The results show that signal intensity can seriously affect the quality of MS / MS spectra, and it may be difficult for analysts to set a suitable MF threshold for MS / MS spectra in complex sample analysis.

[0141] With the help of in-source fragment ions, the false positive results can be reduced to less than 13.3% by setting the matching ion value to 2. Figure 6 It can be seen from 6a, 6b and 6c that the false positive rates are almost the same under different MS / MS spectrum similarity thresholds. Figure 6 Taking the value in a as an example, when the MS / MS spectrum similarity threshold is 0.6, 0.7 and 0.8, the false positive rate is 13.3%. Larger matching ion values, such as 3 or 4, can further reduce false positive results, but unfortunately, this may lead to the problem of false negative results. Figure 6 The results in show that the minimum number of fragment ions for matching MS / MS spectra can be optimized to 2. A significant advantage of the developed strategy is that it is insensitive to the similarity threshold of the MS / MS spectra.

[0142] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for accurately identifying chemical components in complex plant matrices based on in-source pyrolysis multi-fragment information, characterized in that: The following steps are involved: S1. Prepare the sample solution to be tested; S2. Collect sample data; S3. Extract ion chromatograms for peak detection and annotation; S4. Identify the fragment ions in the source; S5. Identify compounds based on in-source fragment ions.

2. The method according to claim 1, characterized in that The specific method of S1 is: The sample to be tested is ultrasonically extracted with a mixture of methanol and water, and then centrifuged to obtain the supernatant.

3. The method according to claim 1, characterized in that The specific method of S2 collecting sample data is: The sample solution was subjected to UHPLC-HRMS analysis, and the compounds were separated using a Waters Acquity UPLC HSS T3 column, 50 mm × 2.1 mm, with a particle size of 1.8 μm; The column temperature was set at 35°C; The mobile phase consisted of two parts: phase A was 0.1% formic acid aqueous solution and organic phase, and phase B was 0.1% formic acid acetonitrile solution; the flow rate was 0.2 mL / min; The elution gradient was set as follows: 0 min, 97% A; 1 min, 97% A; 1.5 min, 85% A; 23 min, 50% A; 28 min, 30% A; 33 min, 2% A; 37 min, 2% A; 37 min, 97% A; 40 min, 97% A; In terms of mass spectrometry parameters, the psi value of GS1 and GS2 was set to 50, and the curtain gas pressure was 30 psi; The ion source temperature was set to 500°C; The values ​​of ISVF and DP were set to 5500 V and 80 V, respectively; Using DDA data acquisition mode, MS 1 The scanning range is 100-1000Da, and the resolution is 15000; The scan range of MS / MS was set to 50–1000 Da; The mass tolerance of the precursor ion was 100 mDa; The collision energy of the product ions was 35 V, and the collision assist voltage was 15 V; The dynamic exclusion time was set to 5 seconds.

4. The method according to claim 1, characterized in that: The S3 is performed in the AntDAS-DDA platform.

5. The method according to claim 1, characterized in that: The S4 method for identifying fragment ions in the source is: i) selecting an EIC peak m having an MS / MS spectrum and extracting ions from the MS / MS spectrum to search for potential in-source fragment ions; ii) if the high-precision m / z value of another EIC peak n matches an ion in the MS / MS spectrum in step i), then use this peak for in-source fragmentation checking; iii) Pearson's correlation coefficient was used to characterize the similarity of peaks m and n, which was calculated from their elution curves; The calculation formula is: Among them, S m,n is the Pearson correlation coefficient of EIC peaks m and n; c m and c n are the chromatogram vectors corresponding to EIC peaks m and n, respectively; and It is c m and c n The superscript T indicates the transpose of the vector; the ions with a Pearson correlation coefficient exceeding 0.975 will be marked as in-source fragment ions from the same compound; iv) Construct MS for each compound based on the identified in-source fragment ions 1 Spectra, and finally the MS of the in-source fragment ions 1 Spectra and MS / MS spectra were used for compound identification.

6. The method according to claim 1, characterized in that The method for identifying compounds based on the fragment ions in the S5 source is: i) introducing the precursor ions into a compound database to search for candidates that meet the m / z tolerance; ii) Compare the in-source fragment ions of the unknown compound with the fragment ions of the candidate in the library. Specifically, compare the MS / MS spectra of the in-source fragment ions with the corresponding spectra in the library to calculate the similarity or match factor MF, as shown below: Among them, MF i is the i-th in-source fragment ion of the unknown compound; p i is the normalized MS / MS spectrum corresponding to the ion; r j is the normalized MS / MS spectrum of the jth in-source fragment ion of the reference compound in the library; p i The precursor ion should be j The precursor ions are the same; MF i Values ​​greater than 0.5 were considered acceptable matches and used in subsequent analyses; iii) If in MS 1 If there are multiple in-source fragment ions in the spectrum, the MS 1 The matching factor MF of the spectrum is marked as MF MS1 , used for compound identification; iv) Construct a total match factor TMF for each candidate compound, calculated according to the following formula: TMF=MF MS1 +ΣMF i (3) Among them, MF MS1 It is MS 1 Matching factor of the spectrum, MF i is the matching factor of the i-th in-source fragment ion, and the candidate compound with the highest TMF will be selected as the identification result; In practical applications, the low content of a compound may reduce the number of in-source fragment ions that can be generated in the high-resolution mass spectrometry DDA mode, and at least two in-source fragment ions should be matched to ensure effective compound identification.

7. The method according to claim 6, characterized in that The compound database is constructed by: The raw data obtained by UHPLC-HRMS analysis of the compound standards were imported into AntDAS for data analysis, including EIC construction, chromatographic peak extraction, chromatographic peak annotation, fragment ion identification, MS1 and MS / MS spectrum construction, to obtain the MS1 and MS / MS spectrum information of the compound standards; After manual confirmation, the Chinese and English names, CAS numbers, molecular masses, SMILES numbers, molecular formulas, precursor ions, collision voltages, instrument information, and mass spectrum information of the compounds were recorded and integrated to complete the construction of the database.

Images