Data processing method for polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives
Through machine learning models, the secondary mass spectra of polycyclic aromatic hydrocarbons and their derivatives in environmental samples are characterized and screened to generate a list of molecular formulas, which solves the problem of low recognition accuracy in the prior art and achieves more efficient qualitative and quantitative analysis.
Patent Information
- Application Number
- CN202411798969.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-09
- Publication Date
- 2025-04-15
AI Technical Summary
In the prior art, the identification accuracy of polycyclic aromatic hydrocarbons (PAHs) and their derivatives in environmental samples is low, affecting the accuracy of qualitative and quantitative analysis.
Machine learning model is used to characterize the secondary mass spectra of analytes, screen the secondary mass spectra of double charge and single charge, and generate a list of molecular formulas for qualitative and quantitative analysis of PAHs and their derivatives.
It improves the identification accuracy and identification coverage of PAHs and their derivatives, reduces the workload of manual verification, and improves the analysis efficiency.
Smart Images
Figure CN119833024A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing technology, and in particular to a method for processing data of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives. Background Art
[0002] Polycyclic Aromatic Hydrocarbons (PAHs) are a class of aromatic compounds containing two or more benzene rings. They are persistent in the environment due to their stable chemical structure. PAHs come from natural sources (such as forest fires) and anthropogenic sources (such as fossil fuel combustion), accumulate in ecosystems and food chains, and pose a serious threat to human health. Therefore, how to qualitatively and quantitatively identify PAHs and PAHs derivatives such as alkyl PAHs, nitro PAHs, halogenated PAHs, and oxy-PAHs has become a topic of concern.
[0003] In the related art, the aforementioned diverse PAHs derivatives in environmental samples can be qualitatively and quantitatively identified by gas chromatography-mass spectrometry (GC-MS), wherein GC-MS usually relies on library search and mass defect analysis. However, the former is limited by the coverage of the library and cannot effectively identify all known and unknown PAHs; the latter, the presence of a large number of non-biological related mass spectral interference signals (such as isotopes, adducts and neutral losses) in the mass spectral data reduces the accuracy of PAHs and PAHs derivatives identification, affecting the precise qualitative and precise quantification of PAHs and PAHs derivatives. Summary of the invention
[0004] The purpose of the embodiments of the present application is to provide a method, device, computer equipment and computer storable medium for processing data of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives, which can solve the problem of low accuracy in identifying PAHs and PAHs derivatives in related technologies, affecting the accurate qualitative and precise quantitative analysis of PAHs and PAHs derivatives.
[0005] In order to solve the above technical problems, this application is implemented as follows:
[0006] In a first aspect, an embodiment of the present application provides a method for processing data of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives, which may include:
[0007] Obtaining N secondary mass spectra of the analyte, where N is an integer greater than 1;
[0008] Through the machine learning model, feature recognition is performed on each of the N secondary mass spectra to obtain feature distribution evaluation results of M features corresponding to each secondary mass spectra, where M is a positive integer;
[0009] Screening a first target secondary mass spectrum from the N secondary mass spectra, wherein the first target secondary mass spectrum is a secondary mass spectrum including a double-charged secondary mass spectrum and a single-charged secondary mass spectrum matching the double-charged secondary mass spectrum among the N secondary mass spectra;
[0010] Based on the feature distribution evaluation results of M features corresponding to each secondary mass spectrum and the first target secondary mass spectrum, a molecular formula list of analytes is generated, wherein the molecular formula list includes at least one molecular formula for representing the composition and structure of the analyte, and the molecular formula is used for qualitative and quantitative analysis of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives.
[0011] In some embodiments provided in the present application, before the above-mentioned step of “obtaining N secondary mass spectra of the analyte”, the method may further include:
[0012] The mass spectrum data of the analyte are deconvoluted and purified to obtain N secondary mass spectra.
[0013] In some embodiments provided in the present application, the above-mentioned step of "performing feature recognition on each of N secondary mass spectra through a machine learning model to obtain a feature distribution evaluation result of M features corresponding to each secondary mass spectra" may specifically include:
[0014] In the case where the machine learning model includes a normalization model and a random forest model, each of the N secondary mass spectra is normalized by the normalization model to obtain a normalized secondary mass spectrum corresponding to each secondary mass spectrum;
[0015] The normalized secondary mass spectrum is subjected to feature recognition through the random forest model, and feature distribution evaluation results of M features corresponding to the normalized secondary mass spectrum are obtained.
[0016] In some embodiments provided in the present application, the above-mentioned step of “normalizing each of the N secondary mass spectra through a normalization model to obtain a normalized secondary mass spectrum corresponding to each secondary mass spectrum” may specifically include:
[0017] In the case where each of the N secondary mass spectra corresponds to a mass-to-charge ratio of a molecular ion peak, the ion summary intensity of all fragments of the N secondary mass spectra is obtained according to the cumulative value of the ion intensity corresponding to each of the N secondary mass spectra;
[0018] The summary mass-to-charge ratio associated with the summary intensity of the ions and the mass-to-charge ratio of the fragments of each secondary mass spectrum are normalized to obtain a normalized secondary mass spectrum corresponding to each secondary mass spectrum.
[0019] In some embodiments provided in the present application, the M features mentioned above include at least one of the following: regional intensity ratio, skewness, entropy, base-peak ratio;
[0020] Among them, the regional intensity ratio is used to characterize the proportion of the ion signal intensity of each secondary mass spectra to the total ion signal intensity of N secondary mass spectra;
[0021] The skewness is used to characterize the symmetry of the mass-to-charge ratio of each secondary mass spectrum among the mass-to-charge ratios of N secondary mass spectrum fragments;
[0022] Entropy is used to characterize the degree of disorder of the mass-to-charge ratio of each secondary mass spectrum among the mass-to-charge ratios of N secondary mass spectrum fragments;
[0023] The base peak ratio is used to characterize the relative degree of the base peak mass-to-charge ratio and the molecular ion peak mass-to-charge ratio in each secondary mass spectrum.
[0024] In some embodiments provided in the present application, the above-mentioned step of “performing feature recognition on the normalized secondary mass spectrum through a random forest model to obtain feature distribution evaluation results of M features corresponding to the normalized secondary mass spectrum” may specifically include:
[0025] When the M features are regional intensity ratios and the feature distribution results include a first feature distribution result corresponding to the regional intensity ratio, the ion signal intensity of the normalized secondary mass spectrometer and the total ion signal intensity of the N normalized secondary mass spectra are identified by a random forest model; wherein the total ion signal intensity of the N normalized secondary mass spectra is the total ion signal intensity of the N secondary mass spectra;
[0026] The ratio of the ion signal intensity of the normalized secondary mass spectrometer to the total ion signal intensity of N normalized secondary mass spectra is determined as the first characteristic distribution result.
[0027] In some embodiments provided in the present application, the above-mentioned step of “performing feature recognition on the normalized secondary mass spectrum through a random forest model to obtain feature distribution evaluation results of M features corresponding to the normalized secondary mass spectrum” may specifically include:
[0028] When the M features are skewness, the feature distribution result includes a second feature distribution result corresponding to the skewness, each of the N secondary mass spectra corresponds to a mass-to-charge ratio, and the mass-to-charge ratio corresponding to the normalized secondary mass spectrum is the mass-to-charge ratio corresponding to the secondary mass spectrum corresponding to the normalized secondary mass spectrum, the difference between the mass-to-charge ratio corresponding to the normalized secondary mass spectrum and the average mass-to-charge ratio of the mass-to-charge ratio corresponding to the normalized secondary mass spectrum is calculated according to the mass-to-charge ratio corresponding to the normalized secondary mass spectrum and the average mass-to-charge ratio of the N normalized secondary mass spectra;
[0029] The ratio of the difference value to the standard deviation of the mass-to-charge ratios corresponding to the N normalized secondary mass spectra and the first quantity product of the normalized secondary mass spectra is determined as the second characteristic distribution result; wherein the first quantity is N-1.
[0030] In some embodiments provided in the present application, the above-mentioned step of “performing feature recognition on the normalized secondary mass spectrum through a random forest model to obtain feature distribution evaluation results of M features corresponding to the normalized secondary mass spectrum” may specifically include:
[0031] When the M features are entropy and the feature distribution result includes a third feature distribution result corresponding to the entropy, identifying relative ion signal intensities of P ion peaks in the normalized secondary mass spectrum, where P is an integer greater than 1;
[0032] According to the relative ion signal intensity of each ion peak in the P ion peaks, the aggregate relative ion signal intensity of the P ion peaks is counted;
[0033] The ion signal intensity after normalization of the summarized relative ion signal intensity is determined as the third characteristic distribution result.
[0034] In some embodiments provided in the present application, the above-mentioned step of “screening a first target secondary mass spectrum from N secondary mass spectra” may specifically include:
[0035] According to the isotope difference value of the mass-to-charge ratio between the doubly charged fragments, a candidate secondary mass spectrum is screened from the N secondary mass spectra, the candidate secondary mass spectrum being a doubly charged secondary mass spectrum;
[0036] According to the first mass-to-charge ratio corresponding to the candidate secondary mass spectrum, a second mass-to-charge ratio corresponding to the first mass-to-charge ratio is calculated, wherein the second mass-to-charge ratio is the mass-to-charge ratio corresponding to the single-charged secondary mass spectrum matching the candidate secondary mass spectrum;
[0037] Based on the second mass-to-charge ratio, screening a single-charged secondary mass spectrum matching the candidate secondary mass spectrum from the N secondary mass spectra;
[0038] The singly charged secondary mass spectrum and the candidate secondary mass spectrum are determined as the first target secondary mass spectrum.
[0039] In some embodiments provided in the present application, the above-mentioned step of "generating a molecular formula list of analytes according to the feature distribution evaluation results of M features corresponding to each secondary mass spectrum and the first target secondary mass spectrum" may specifically include:
[0040] According to the characteristic distribution evaluation results of the M features corresponding to each secondary mass spectrum, a second target secondary mass spectrum is screened from the N secondary mass spectra, and the characteristic distribution evaluation results of the M features corresponding to the second target secondary mass spectrum meet the preset conditions;
[0041] Molecular formula prediction is performed on the first target secondary mass spectrum and the second target secondary mass spectrum to obtain a molecular formula list of the analyte.
[0042] In a second aspect, an embodiment of the present application provides a device for processing data of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives, which may include:
[0043] An acquisition module, used for acquiring N secondary mass spectra of the analyte, where N is an integer greater than 1;
[0044] An identification module is used to perform feature identification on each of the N secondary mass spectra through a machine learning model to obtain a feature distribution evaluation result of M features corresponding to each secondary mass spectra, where M is a positive integer;
[0045] A screening module, used for screening a first target secondary mass spectrum from N secondary mass spectra, wherein the first target secondary mass spectrum is a secondary mass spectrum including a double-charged secondary mass spectrum and a single-charged secondary mass spectrum matching the double-charged secondary mass spectrum in the N secondary mass spectra;
[0046] A generation module is used to generate a molecular formula list of analytes based on the feature distribution evaluation results of M features corresponding to each secondary mass spectrum and the first target secondary mass spectrum, wherein the molecular formula list includes at least one molecular formula for representing the composition and structure of the analyte, and the molecular formula is used for qualitative and quantitative analysis of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives.
[0047] In some embodiments provided in the present application, the above-mentioned recognition module can also be used to deconvolute and purify the mass spectrum data of the analyte to obtain N secondary mass spectra.
[0048] In some embodiments provided in the present application, the above-mentioned polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbons derivatives data processing device may further include a processing module; wherein,
[0049] A processing module, for performing normalization processing on each of the N secondary mass spectra by using the normalization model to obtain a normalized secondary mass spectrum corresponding to each secondary mass spectrum when the machine learning model includes a normalization model and a random forest model;
[0050] The recognition module can also be used to perform feature recognition on the normalized secondary mass spectrum through a random forest model to obtain feature distribution evaluation results of M features corresponding to the normalized secondary mass spectrum.
[0051] In some embodiments provided in the present application, the above-mentioned processing module may also be used to obtain the ion summary intensity of all fragments of the N secondary mass spectra according to the cumulative value of the ion intensity corresponding to each secondary mass spectrum in the N secondary mass spectra, when each secondary mass spectrum in the N secondary mass spectra corresponds to the mass-to-charge ratio of a molecular ion peak;
[0052] The summary mass-to-charge ratio associated with the summary intensity of the ions and the mass-to-charge ratio of the fragments of each secondary mass spectrum are normalized to obtain a normalized secondary mass spectrum corresponding to each secondary mass spectrum.
[0053] In some embodiments provided in the present application, the M features mentioned above include at least one of the following: regional intensity ratio, skewness, entropy, base-peak ratio;
[0054] Among them, the regional intensity ratio is used to characterize the proportion of the ion signal intensity of each secondary mass spectra to the total ion signal intensity of N secondary mass spectra;
[0055] The skewness is used to characterize the symmetry of the mass-to-charge ratio of each secondary mass spectrum among the mass-to-charge ratios of N secondary mass spectra;
[0056] Entropy is used to characterize the degree of disorder of the mass-to-charge ratio of each secondary mass spectrum among the mass-to-charge ratios of N secondary mass spectra;
[0057] The base peak ratio is used to characterize the relative degree of the base peak mass-to-charge ratio and the molecular ion peak mass-to-charge ratio in each secondary mass spectrum.
[0058] In some embodiments provided in the present application, the above-mentioned identification module can be specifically used to identify the ion signal intensity of the normalized secondary mass spectrometer and the total ion signal intensity of N normalized secondary mass spectra through a random forest model when M features are regional intensity ratios and the feature distribution result includes a first feature distribution result corresponding to the regional intensity ratio; wherein the total ion signal intensity of the N normalized secondary mass spectra is the total ion signal intensity of the N secondary mass spectra;
[0059] The ratio of the ion signal intensity of the normalized secondary mass spectrometer to the total ion signal intensity of N normalized secondary mass spectra is determined as the first characteristic distribution result.
[0060] In some embodiments provided in the present application, the above-mentioned identification module can be specifically used for, when M features are skewness, the feature distribution result includes a second feature distribution result corresponding to the skewness, each secondary mass spectrum in N secondary mass spectra corresponds to a mass-to-charge ratio, and the mass-to-charge ratio corresponding to the normalized secondary mass spectrum is the mass-to-charge ratio corresponding to the secondary mass spectrum corresponding to the normalized secondary mass spectrum, according to the mass-to-charge ratio corresponding to the normalized secondary mass spectrum and the average value of the mass-to-charge ratio corresponding to the N normalized secondary mass spectra, calculate the difference between the mass-to-charge ratio corresponding to the normalized secondary mass spectrum and the average mass-to-charge ratio;
[0061] The ratio of the difference value to the standard deviation of the mass-to-charge ratios corresponding to the N normalized secondary mass spectra and the first quantity product of the normalized secondary mass spectra is determined as the second characteristic distribution result; wherein the first quantity is N-1.
[0062] In some embodiments provided in the present application, the above-mentioned identification module can be specifically used to identify the relative ion signal intensities of P ion peaks in the normalized secondary mass spectrum when M features are entropy and the feature distribution result includes a third feature distribution result corresponding to the entropy, where P is an integer greater than 1;
[0063] According to the relative ion signal intensity of each ion peak in the P ion peaks, the aggregate relative ion signal intensity of the P ion peaks is counted;
[0064] The ion signal intensity after normalization of the summarized relative ion signal intensity is determined as the third characteristic distribution result.
[0065] In some embodiments provided in the present application, the above-mentioned screening module can also be used to screen a candidate secondary mass spectrum from N secondary mass spectra according to the isotope difference value of the mass-to-charge ratio between the doubly charged fragments, wherein the candidate secondary mass spectrum is a doubly charged secondary mass spectrum;
[0066] According to the first mass-to-charge ratio corresponding to the candidate secondary mass spectrum, a second mass-to-charge ratio corresponding to the first mass-to-charge ratio is calculated, wherein the second mass-to-charge ratio is the mass-to-charge ratio corresponding to the single-charged secondary mass spectrum matching the candidate secondary mass spectrum;
[0067] Based on the second mass-to-charge ratio, screening a single-charged secondary mass spectrum matching the candidate secondary mass spectrum from the N secondary mass spectra;
[0068] The singly charged secondary mass spectrum and the candidate secondary mass spectrum are determined as the first target secondary mass spectrum.
[0069] In some embodiments provided in the present application, the above-mentioned generation module may also be applied to screen a second target secondary mass spectrum from N secondary mass spectra according to the feature distribution evaluation results of the M features corresponding to each secondary mass spectrum, and the feature distribution evaluation results of the M features corresponding to the second target secondary mass spectrum meet the preset conditions;
[0070] Molecular formula prediction is performed on the first target secondary mass spectrum and the second target secondary mass spectrum to obtain a molecular formula list of the analyte.
[0071] In a third aspect, an embodiment of the present application provides a computer device, which includes a processor, a memory, and a program or instruction stored in the memory and executable on the processor, and when the program or instruction is executed by the processor, the steps of the polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives data processing method shown in the first aspect are implemented.
[0072] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives data processing method shown in the first aspect are implemented.
[0073] In the fifth aspect, an embodiment of the present application provides a chip, which includes a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run programs or instructions to implement the steps of the polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives data processing method shown in the first aspect.
[0074] In summary, the data processing method for polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives provided in the embodiment of the present application, through a machine learning model, performs feature recognition on each secondary mass spectrum of N secondary mass spectra of the analyte, and obtains a feature distribution evaluation result corresponding to M features of each secondary mass spectrum; then, a first target secondary mass spectrum is screened from the N secondary mass spectra, and the first target secondary mass spectrum is a secondary mass spectrum including a double-charged secondary mass spectrum and a single-charged secondary mass spectrum matching the double charge in the N secondary mass spectra; then, according to the feature distribution evaluation result corresponding to the M features of each secondary mass spectrum and the first target secondary mass spectrum, a molecular formula list of the analyte is generated, and the molecular formula list includes at least one molecular formula for representing the composition and structure of the analyte, and the molecular formula is used for qualitative and quantitative analysis of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives. In this way, by introducing the recognition method of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives of machine learning models and high-resolution mass spectrometry technology, the accuracy and identification coverage of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives can be effectively improved. In addition, by introducing the recognition of doubly charged secondary mass spectra and molecular formula prediction, the identification process of PAHs and PAH derivatives was further optimized, and the reliability of the identification results was improved. At the same time, the automated identification workflow of PAHs and PAH derivatives significantly reduced the workload of manual verification and improved the efficiency of identifying PAHs and PAH derivatives. BRIEF DESCRIPTION OF THE DRAWINGS
[0075] Figure 1 A flowchart of a method for processing data of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives provided in an embodiment of the present application;
[0076] Figure 2 A schematic diagram of determining a first target secondary mass spectrum of a method for processing data of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives provided in an embodiment of the present application;
[0077] Figure 3 A schematic diagram of the structure of a data processing device for cyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives provided in an embodiment of the present application;
[0078] Figure 4 A schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0079] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0080] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.
[0081] PAHs are carcinogenic, mutagenic and teratogenic, and can cause diseases such as lung cancer, skin cancer and bladder cancer. In addition, PAHs can cause respiratory and cardiovascular diseases and aggravate symptoms such as asthma. The complex structure of PAHs and the diversity of their derivatives, such as alkyl PAHs, nitro PAHs, halogenated PAHs and oxo PAHs, pose significant challenges. These derivatives exhibit different chemical behaviors and toxicity. For example, nitro PAHs are formed in diesel exhaust and are particularly mutagenic and carcinogenic. Advanced analytical techniques are essential to accurately distinguish and quantify these diverse PAHs derivatives in environmental samples.
[0082] Mass spectrometry is an important technical means in omics research. It analyzes the compounds in the sample by measuring the mass-to-charge ratio (m / z) of molecular ions. The basic principle of mass spectrometry is to ionize the compounds in the sample to generate charged ions. These ions are separated and detected in an electric or magnetic field according to their mass-to-charge ratio. The signal output by the mass spectrometer is usually expressed in the form of a secondary mass spectrum, with the horizontal axis representing the mass-to-charge ratio and the vertical axis representing the relative abundance of the ions.
[0083] In related technologies, GC-MS identification and recognition of PAHs mainly rely on two methods: library search and mass defect analysis.
[0084] Specifically, library retrieval refers to comparing the experimentally obtained mass spectrum with a database of standard mass spectra of known compounds. Unknown compounds are identified by comparing the mass-to-charge ratio (m / z) and the relative abundance of fragment ions. This method relies on high-quality standard libraries, such as the NIST mass spectral library or MassBank. If the PAHs in the sample have a high degree of match with the standard spectra in the database, these compounds can be identified with high confidence. However, the limitation of library retrieval is that the coverage of the library is seriously insufficient. Although there are many mass spectra of standard substances in the mass spectral library, it is still far from covering all known and unknown PAHs compounds. Therefore, when the PAHs in the sample are not in the database, PAHs and PAHs derivatives cannot be accurately identified.
[0085] Mass defect analysis is an analytical method based on the accurate mass of compounds. It normalizes the masses of compounds of different chain lengths by recalibrating the experimental mass spectrometry data to the Kendrick Mass (KM) mass. Mass defect analysis can effectively cluster and identify compounds containing different heteroatoms, which helps to screen potential PAHs candidates. However, the limitation of mass defect analysis is that the number of false positive candidates is very large: the candidate list generated by mass defect analysis can be very large, especially when dealing with complex environmental samples, which requires a lot of subsequent verification work.
[0086] Based on this, in order to solve the above-mentioned problems. The data processing method for polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives provided in the embodiment of the present application is a spectral feature recognition process for PAHs, which aims to extract PAH compounds from non-targeted analysis of gas chromatography high-resolution mass spectrometry (GC-HRMS). The process first deconvolutes and purifies the collected data, and then evaluates the characteristic distribution of N secondary mass spectra of the analyte through a machine learning model. In addition, by identifying doubly charged ions and molecular formula predictions, a potential list of molecular formulas is finally output to identify the structures of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives.
[0087] The following is combined with Figure 1-Figure 2 , the data processing of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives provided in the embodiments of the present application is described in detail through specific examples.
[0088] First, combine Figure 1 The data processing method of PAHs and PAHs derivatives is explained in detail.
[0089] Figure 1 A flowchart of a method for processing data of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives provided in an embodiment of the present application.
[0090] like Figure 1 As shown, the method for processing data of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives may include steps 110 to 140 .
[0091] Step 110, obtaining N secondary mass spectra of the analyte, where N is an integer greater than 1; step 120, performing feature recognition on each of the N secondary mass spectra through a machine learning model to obtain a feature distribution evaluation result of M features corresponding to each secondary mass spectra, where M is a positive integer; step 130, screening a first target secondary mass spectrum from the N secondary mass spectra, where the first target secondary mass spectrum is a secondary mass spectrum including a double-charged secondary mass spectrum and a single-charged secondary mass spectrum matching the double-charged secondary mass spectrum among the N secondary mass spectra; step 140, generating a molecular formula list of the analyte based on the feature distribution evaluation result of the M features corresponding to each secondary mass spectrum and the first target secondary mass spectrum, where the molecular formula list includes at least one molecular formula for representing the composition and structure of the analyte, and the molecular formula is used for qualitative and quantitative analysis of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives.
[0092] In this way, by introducing the recognition method of PAHs and PAHs derivatives using machine learning models and high-resolution mass spectrometry technology, the accuracy and identification coverage of PAHs and PAHs derivatives can be effectively improved. In addition, by introducing the recognition of doubly charged secondary mass spectra and molecular formula prediction, the recognition process of PAHs and PAHs derivatives is further optimized, and the reliability of the identification results is improved. At the same time, the workflow of the automated identification of PAHs and PAHs derivatives significantly reduces the workload of manual verification and improves the efficiency of identifying PAHs and PAHs derivatives.
[0093] The above steps are described in detail below, as shown below.
[0094] First, referring to step 110, in some embodiments of the present application, before step 110, the method for processing data of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives may further include step 150, performing deconvolution and purification on the mass spectrum data of the analyte to obtain N secondary mass spectra.
[0095] Next, step 120 is involved. In some embodiments of the present application, the machine learning model includes a normalization model and a random forest model. Based on this, step 120 may specifically include step 1201 and step 1202, as shown below.
[0096] Step 1201, normalizing each of the N secondary mass spectra using a normalization model to obtain a normalized secondary mass spectrum corresponding to each secondary mass spectrum.
[0097] Step 1202, performing feature recognition on the normalized secondary mass spectrum through a random forest model, and obtaining feature distribution evaluation results of M features corresponding to the normalized secondary mass spectrum.
[0098] Exemplarily, the m / z of all N secondary mass spectra are divided by the m / z of the molecular ion peak of the corresponding secondary mass spectra and normalized to a range of 0-100%. After normalization, the distribution of any secondary mass spectra is comparable, thereby providing a unified framework for analyzing secondary mass spectra.
[0099] Further, in some embodiments, each of the N secondary mass spectra corresponds to the mass-to-charge ratio of a molecular ion peak. Based on this, step 1201 may specifically include step 12011 and step 12012.
[0100] Step 12011, according to the accumulated value of the ion intensity corresponding to each of the N secondary mass spectra, obtain the summary ion intensity of all fragments of the N secondary mass spectra.
[0101] Step 12012, normalize the aggregated mass-to-charge ratio related to the aggregated ion intensity and the mass-to-charge ratio of the fragments of each secondary mass spectrum to obtain a normalized secondary mass spectrum corresponding to each secondary mass spectrum.
[0102] In some embodiments, involving the above step 1202, based on the evaluation framework of the characteristic distribution evaluation results of the M characteristics of the secondary mass spectrum, the embodiment of the present application provides three types of machine learning features to evaluate the characteristics of the spectrum, and trains and optimizes a random forest model to identify the fragment distribution characteristics of PAHs. The random forest model identifies PAHs based on characteristics such as regional intensity ratio, skewness, entropy, and base peak ratio.
[0103] The M features in the embodiment of the present application include at least one of the following: regional intensity ratio, skewness, entropy, and base-peak ratio. Among them, the regional intensity ratio is used to characterize the proportion of the ion signal intensity of each secondary mass spectrogram in the total ion signal intensity of N secondary mass spectrograms. Skewness is used to characterize the degree of symmetry of the distribution of the mass-to-charge ratio of each secondary mass spectrogram in the mass-to-charge ratio of N secondary mass spectrograms. Entropy is used to characterize the degree of disorder of the mass-to-charge ratio of each secondary mass spectrogram in the mass-to-charge ratio of N secondary mass spectrograms. The base-peak ratio is used to characterize the relative degree of the base peak mass-to-charge ratio and the molecular ion peak mass-to-charge ratio in each secondary mass spectrogram.
[0104] Thus, after the m / z is normalized to the range of 0-100% after being divided by the m / z of [M]+ (molecular ion peak), and after being divided into three equal parts, three regional intensity intervals can be obtained. At this time, the difference in the distribution ratio of the intensity ratio values of PAHs and PAHs compounds in these three regions, and the unique regional intensity ratio of PAHs relative to other compounds enable machine learning to learn the spectral features of PAHs for specific identification. Therefore, the above four features can better calculate and evaluate the regional intensity ratio, skewness, entropy, base peak ratio and other features of the fragments in the mass spectrometry data, thereby improving the recognition accuracy and efficiency of polycyclic aromatic hydrocarbons and their derivatives.
[0105] The following describes in detail the feature distribution evaluation results of M features corresponding to the normalized secondary mass spectrum based on different features.
[0106] In some embodiments, the M features are region intensity ratios (RIRs), and the feature distribution results include a first feature distribution result corresponding to the region intensity ratios.
[0107] Based on this, step 1202 may specifically include step 12021 and step 12022.
[0108] Step 12021, using a random forest model, identifies the ion signal intensity of the normalized secondary mass spectrometer and the total ion signal intensity of N normalized secondary mass spectra; wherein the total ion signal intensity of the N normalized secondary mass spectra is the total ion signal intensity of the N secondary mass spectra.
[0109] Step 12022, determine the ratio of the ion signal intensity of the normalized secondary mass spectrometer to the total ion signal intensity of N normalized secondary mass spectra as the first characteristic distribution result.
[0110] For example, RIR can be calculated by the following formula (1):
[0111]
[0112] in, It represents the sum of all fragment intensities in a specific area, that is, the ion signal intensity of the normalized secondary mass spectrum. Represents the total ion signal intensity.
[0113] In other embodiments, the M features are skewness, the feature distribution result includes a second feature distribution result corresponding to the skewness, each of the N secondary mass spectra corresponds to a mass-to-charge ratio, and the mass-to-charge ratio corresponding to the normalized secondary mass spectrum is the mass-to-charge ratio corresponding to the secondary mass spectrum corresponding to the normalized secondary mass spectrum.
[0114] Based on this, step 1202 may specifically include step 12023 and step 12024.
[0115] Step 12023, based on the mass-to-charge ratio corresponding to the normalized secondary mass spectrum and the average mass-to-charge ratio corresponding to N normalized secondary mass spectra, calculate the difference between the mass-to-charge ratio corresponding to the normalized secondary mass spectrum and the average mass-to-charge ratio.
[0116] Step 12024, determine the ratio of the difference value to the product of the standard deviation of the mass-to-charge ratios corresponding to N normalized secondary mass spectra and the first quantity of the normalized secondary mass spectra as the second characteristic distribution result; wherein the first quantity is N-1.
[0117] For example, Skewness can be calculated by the following formula (2):
[0118]
[0119] Where x represents the mass-to-charge ratio corresponding to the normalized secondary mass spectrum, represents the average value of the mass-to-charge ratios corresponding to N normalized secondary mass spectra, N represents the second number of normalized secondary mass spectra, and the value of the second number is N, and s represents the standard deviation value of the mass-to-charge ratios corresponding to the N normalized secondary mass spectra.
[0120] In some further embodiments, the M features are entropy, and the feature distribution result includes a third feature distribution result corresponding to the entropy.
[0121] Based on this, step 1202 may specifically include step 12025 and step 12027.
[0122] Step 12025, identifying the relative ion signal intensities of P ion peaks in the normalized secondary mass spectrum, where P is an integer greater than 1.
[0123] Step 12026, based on the relative ion signal intensity of each ion peak in the P ion peaks, the aggregate relative ion signal intensity of the P ion peaks is counted.
[0124] Step 12027, determining the ion signal intensity after normalization of the aggregated relative ion signal intensity as the third characteristic distribution result.
[0125] For example, Entropy can be calculated by the following formula (3):
[0126]
[0127] Where x represents the relative ion signal intensity of each ion peak in the normalized secondary mass spectrum, N represents the total number of normalized secondary mass spectra, and log represents logarithmic operation. The summary relative ion signal intensities were normalized to convert them into probabilities.
[0128] In some further embodiments, the M features are base peak ratios (BPR), and the feature distribution results include a fourth feature distribution result corresponding to the base peak ratio, and the BPR is used to characterize the relative degree of the base peak mass-to-charge ratio and the molecular ion peak mass-to-charge ratio in each secondary mass spectrum.
[0129] Based on this, step 1202 may specifically include step 12028.
[0130] Step 12028, the ratio of the mass-to-charge ratio of the base peak fragment in the normalized secondary mass spectrum to the mass-to-charge ratio of the molecular ion peaks in the N normalized secondary mass spectra is determined as the base peak ratio. The base peak fragment can be the peak with the highest ion signal intensity in the normalized secondary mass spectrum.
[0131] For example, the BPR can be calculated by the following formula (4):
[0132]
[0133] Among them, x B represents the mass-to-charge ratio of the base peak fragment (i.e., the peak with the highest intensity) in the normalized secondary mass spectrum, x M Represents the mass-to-charge ratio of the molecular ion peaks of N normalized secondary mass spectra.
[0134] Therefore, the data processing method of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives in the embodiment of the present application can improve the identification accuracy and efficiency of polycyclic aromatic hydrocarbons and their derivatives by calculating and evaluating the characteristics such as regional intensity ratio, skewness and entropy of the fragments in the mass spectrometry data.
[0135] Furthermore, with respect to step 130, in some embodiments of the present application, step 130 may specifically include steps 1301 to 1304, as shown below.
[0136] Step 1301, screening a candidate secondary mass spectrum from N secondary mass spectra according to the isotope difference value of the mass-to-charge ratio between the doubly charged fragments, the candidate secondary mass spectrum being a doubly charged secondary mass spectrum.
[0137] Step 1302, based on the first mass-to-charge ratio corresponding to the candidate secondary mass spectrum, calculate a second mass-to-charge ratio corresponding to the first mass-to-charge ratio, the second mass-to-charge ratio being the mass-to-charge ratio corresponding to the single-charged secondary mass spectrum matching the candidate secondary mass spectrum.
[0138] Step 1303: Based on the second mass-to-charge ratio, a singly charged secondary mass spectrum matching the candidate secondary mass spectrum is screened from the N secondary mass spectra.
[0139] Step 1304: determine the singly charged secondary mass spectrum and the candidate secondary mass spectrum as the first target secondary mass spectrum.
[0140] For example, Figure 2 As shown, by calculating the m / z-specific isotopic difference between the doubly charged fragments, that is, the isotopic difference value of the mass-to-charge ratio between the doubly charged fragments, such as 0.5 Dalton, the doubly charged secondary mass spectra are identified, such as the secondary mass spectra corresponding to the first mass-to-charge ratios of 276, 276.5, and 277. Then, for each candidate doubly charged secondary mass spectra, the expected mass-to-charge ratio corresponding to the matching single-charged secondary mass spectra is calculated, for example, the first mass-to-charge ratio is 276, and the second mass-to-charge ratio corresponding to the first mass-to-charge ratio 276 is 138. At this time, the single-charged secondary mass spectra with the second mass-to-charge ratio of 138 and the doubly charged secondary mass spectra with the first mass-to-charge ratio of 276 can be determined as the first target secondary mass spectra.
[0141] Therefore, doubly charged fragments can be identified by high-resolution mass spectrometry based on the m / z differences of isotopes between fragments, and combined with molecular formula prediction, the accuracy and reliability of the identification results can be ensured, effectively solving the problem of PAHs identification in complex samples.
[0142] Then, step 140 is involved. In some embodiments of the present application, step 140 may specifically include step 1401 and step 1402, as shown below.
[0143] Step 1401, based on the feature distribution evaluation results of M features corresponding to each secondary mass spectrum, a second target secondary mass spectrum is screened from N secondary mass spectra, and the feature distribution evaluation results of the M features corresponding to the second target secondary mass spectrum meet preset conditions.
[0144] Step 1402, performing molecular formula prediction on the first target secondary mass spectrum and the second target secondary mass spectrum to obtain a molecular formula list of the analyte.
[0145] For example, after confirming the doubly charged fragments, the Rdisop tool can be used to predict the molecular formula, limiting the elemental composition within a certain range (the carbon-hydrogen ratio of the PAHs molecular formula is significantly higher than that of other types of molecules). The accuracy of PAHs identification is ensured through molecular formula verification. The figure above shows that doubly charged fragments are a mass spectral feature unique to PAHs under high-resolution mass spectrometry.
[0146] Therefore, the automated PAHs identification workflow can significantly reduce the workload of manual intervention through standardized and automated processing steps, improve the efficiency and accuracy of PAHs identification, and is suitable for rapid screening and identification of large-scale environmental samples.
[0147] Based on the same inventive concept, the present application provides a data processing device for polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives, specifically in combination with Figure 3 Provide detailed explanation.
[0148] Figure 3 A schematic diagram of the structure of a polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbons derivatives data processing device provided in an embodiment of the present application.
[0149] like Figure 3 As shown, the polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbons derivatives data processing device 30 is applied to a computer device, and specifically may include:
[0150] An acquisition module 301 is used to acquire N secondary mass spectra of the analyte, where N is an integer greater than 1;
[0151] The identification module 302 is used to perform feature identification on each of the N secondary mass spectra through a machine learning model to obtain a feature distribution evaluation result of M features corresponding to each secondary mass spectra, where M is a positive integer;
[0152] A screening module 303 is used to screen a first target secondary mass spectrum from the N secondary mass spectra, where the first target secondary mass spectrum is a secondary mass spectrum including a double-charged secondary mass spectrum and a single-charged secondary mass spectrum matching the double-charged secondary mass spectrum among the N secondary mass spectra;
[0153] A generation module 304 is used to generate a molecular formula list of analytes based on the feature distribution evaluation results of M features corresponding to each secondary mass spectrum and the first target secondary mass spectrum, wherein the molecular formula list includes at least one molecular formula for representing the composition and structure of the analyte, and the molecular formula is used for qualitative and quantitative analysis of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives.
[0154] The polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbons derivatives data processing device 30 provided in the embodiments of the present application is described in detail below.
[0155] In some embodiments provided in the present application, the above-mentioned identification module 302 can also be used to deconvolute and purify the mass spectrum data of the analyte to obtain N secondary mass spectra.
[0156] In some embodiments provided in the present application, the above-mentioned polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbons derivatives data processing device may further include a processing module; wherein,
[0157] A processing module, for performing normalization processing on each of the N secondary mass spectra by using the normalization model to obtain a normalized secondary mass spectrum corresponding to each secondary mass spectrum when the machine learning model includes a normalization model and a random forest model;
[0158] The identification module 302 can also be used to perform feature identification on the normalized secondary mass spectrum through a random forest model to obtain feature distribution evaluation results of M features corresponding to the normalized secondary mass spectrum.
[0159] In some embodiments provided in the present application, the above-mentioned processing module may also be used to obtain the ion summary intensity of all fragments of the N secondary mass spectra according to the cumulative value of the ion intensity corresponding to each secondary mass spectrum in the N secondary mass spectra, when each secondary mass spectrum in the N secondary mass spectra corresponds to the mass-to-charge ratio of a molecular ion peak;
[0160] The summary mass-to-charge ratio associated with the summary intensity of the ions and the mass-to-charge ratio of the fragments of each secondary mass spectrum are normalized to obtain a normalized secondary mass spectrum corresponding to each secondary mass spectrum.
[0161] In some embodiments provided in the present application, the M features mentioned above include at least one of the following: regional intensity ratio, skewness, entropy, base-peak ratio;
[0162] Among them, the regional intensity ratio is used to characterize the proportion of the ion signal intensity of each secondary mass spectra to the total ion signal intensity of N secondary mass spectra;
[0163] The skewness is used to characterize the symmetry of the mass-to-charge ratio of each secondary mass spectrum among the mass-to-charge ratios of N secondary mass spectra;
[0164] Entropy is used to characterize the degree of disorder of the mass-to-charge ratio of each secondary mass spectrum among the mass-to-charge ratios of N secondary mass spectra;
[0165] The base peak ratio is used to characterize the relative degree of the base peak mass-to-charge ratio and the molecular ion peak mass-to-charge ratio in each secondary mass spectrum.
[0166] In some embodiments provided in the present application, the identification module 302 involved above can be specifically used to, when M features are regional intensity ratios and the feature distribution result includes a first feature distribution result corresponding to the regional intensity ratio, identify the ion signal intensity of the normalized secondary mass spectrometer and the total ion signal intensity of N normalized secondary mass spectra through a random forest model; wherein the total ion signal intensity of the N normalized secondary mass spectra is the total ion signal intensity of the N secondary mass spectra;
[0167] The ratio of the ion signal intensity of the normalized secondary mass spectrometer to the total ion signal intensity of N normalized secondary mass spectra is determined as the first characteristic distribution result.
[0168] In some embodiments provided in the present application, the identification module 302 involved above can be specifically used for, when M features are skewness, the feature distribution result includes a second feature distribution result corresponding to the skewness, each secondary mass spectrum in N secondary mass spectra corresponds to a mass-to-charge ratio, and the mass-to-charge ratio corresponding to the normalized secondary mass spectrum is the mass-to-charge ratio corresponding to the secondary mass spectrum corresponding to the normalized secondary mass spectrum, according to the mass-to-charge ratio corresponding to the normalized secondary mass spectrum and the average value of the mass-to-charge ratio corresponding to the N normalized secondary mass spectra, to calculate the difference between the mass-to-charge ratio corresponding to the normalized secondary mass spectrum and the average mass-to-charge ratio;
[0169] The ratio of the difference value to the standard deviation of the mass-to-charge ratios corresponding to the N normalized secondary mass spectra and the first quantity product of the normalized secondary mass spectra is determined as the second characteristic distribution result; wherein the first quantity is N-1.
[0170] In some embodiments provided in the present application, the identification module 302 involved above can be specifically used to identify the relative ion signal intensities of P ion peaks in the normalized secondary mass spectrum when M features are entropy and the feature distribution result includes a third feature distribution result corresponding to the entropy, where P is an integer greater than 1;
[0171] According to the relative ion signal intensity of each ion peak in the P ion peaks, the aggregate relative ion signal intensity of the P ion peaks is counted;
[0172] The ion signal intensity after normalization of the summarized relative ion signal intensity is determined as the third characteristic distribution result.
[0173] In some embodiments provided in the present application, the screening module 303 involved above can also be used to screen a candidate secondary mass spectrum from N secondary mass spectra according to the isotope difference value of the mass-to-charge ratio between the doubly charged fragments, wherein the candidate secondary mass spectrum is a doubly charged secondary mass spectrum;
[0174] According to the first mass-to-charge ratio corresponding to the candidate secondary mass spectrum, a second mass-to-charge ratio corresponding to the first mass-to-charge ratio is calculated, wherein the second mass-to-charge ratio is the mass-to-charge ratio corresponding to the single-charged secondary mass spectrum matching the candidate secondary mass spectrum;
[0175] Based on the second mass-to-charge ratio, screening a single-charged secondary mass spectrum matching the candidate secondary mass spectrum from the N secondary mass spectra;
[0176] The singly charged secondary mass spectrum and the candidate secondary mass spectrum are determined as the first target secondary mass spectrum.
[0177] In some embodiments provided in the present application, the generation module 304 involved above can also be applied to screen a second target secondary mass spectrum from N secondary mass spectra according to the feature distribution evaluation results of the M features corresponding to each secondary mass spectrum, and the feature distribution evaluation results of the M features corresponding to the second target secondary mass spectrum meet the preset conditions;
[0178] Molecular formula prediction is performed on the first target secondary mass spectrum and the second target secondary mass spectrum to obtain a molecular formula list of the analyte.
[0179] The PAH and PAH derivative data processing device in the embodiment of the present application can be a device, or a component, machine circuit, or chip in a computer device. The device can be a mobile computer device, or a non-mobile computer device.
[0180] Exemplarily, the mobile computer device may be a mobile phone, a tablet computer, a laptop computer, a PDA, a vehicle-mounted computer device, a wearable device, an ultra-mobile personal computer (UMPC), a netbook, or a personal digital assistant (PDA), etc., and the non-mobile computer device may be a server, a network attached storage (NAS), a personal computer (PC), a television (TV), a teller machine, or a self-service machine, etc., which is not specifically limited in the embodiments of the present application.
[0181] The operating system of the polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives data processing device in the embodiment of the present application can be an Android operating system, an IOS operating system, or other possible operating systems, which are not specifically limited in the embodiment of the present application.
[0182] The polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbons derivatives data processing device provided in the embodiment of the present application can achieve Figure 1 to Figure 2 To avoid repetition, the various processes implemented by the method embodiment are not described here.
[0183] In summary, the data processing device for polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives provided in the embodiment of the present application performs feature recognition on each of the N secondary mass spectra of the analyte through a machine learning model, and obtains a feature distribution evaluation result corresponding to M features of each secondary mass spectra; then, a first target secondary mass spectrum is screened from the N secondary mass spectra, and the first target secondary mass spectrum is a secondary mass spectrum including a double-charged secondary mass spectrum and a single-charged secondary mass spectrum matching the double charge in the N secondary mass spectra; then, according to the feature distribution evaluation result corresponding to the M features of each secondary mass spectrum and the first target secondary mass spectrum, a molecular formula list of the analyte is generated, and the molecular formula list includes at least one molecular formula for representing the composition and structure of the analyte, and the molecular formula is used for qualitative and quantitative analysis of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives. In this way, by introducing the recognition method of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives using machine learning models and high-resolution mass spectrometry technology, the accuracy and identification coverage of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives can be effectively improved. In addition, by introducing the recognition of doubly charged secondary mass spectra and molecular formula prediction, the identification process of PAHs and PAH derivatives was further optimized, and the reliability of the identification results was improved. At the same time, the automated identification workflow of PAHs and PAH derivatives significantly reduced the workload of manual verification and improved the efficiency of identifying PAHs and PAH derivatives.
[0184] Optional, such as Figure 4 As shown, the embodiment of the present application also provides a computer device 400, including a processor 401, a memory 402, and a program or instruction stored in the memory 402 and executable on the processor 401. When the program or instruction is executed by the processor 401, each process of the above-mentioned polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives data processing method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.
[0185] An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, each process of the above-mentioned polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives data processing method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.
[0186] The processor is the processor in the computer device in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.
[0187] In addition, an embodiment of the present application further provides a chip, which includes a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives data processing method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.
[0188] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.
[0189] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.
[0190] In addition, it should be noted that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in a reverse order according to the functions involved. For example, the described methods may be performed in an order different from that described, and various steps may be added, omitted, or combined. In addition, features described with reference to certain examples may be combined in other examples.
[0191] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a disk, or an optical disk), and includes a number of instructions for a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods of each embodiment of the present application.
[0192] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.
Claims
1. A method for processing data of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives, characterized in that: include: Obtaining N secondary mass spectra of the analyte, where N is an integer greater than 1; Performing feature recognition on each of the N secondary mass spectra through a machine learning model to obtain feature distribution evaluation results of M features corresponding to each secondary mass spectra, where M is a positive integer; Screening a first target secondary mass spectrum from the N secondary mass spectra, wherein the first target secondary mass spectrum is a secondary mass spectrum including a double charge and a single charge matching the double charge among the N secondary mass spectra; Based on the feature distribution evaluation results of the M features corresponding to each secondary mass spectrum and the first target secondary mass spectrum, a molecular formula list of the analyte is generated, wherein the molecular formula list includes at least one molecular formula for representing the composition and structure of the analyte, and the molecular formula is used for qualitative and quantitative analysis of polycyclic aromatic hydrocarbons and polycyclic aromatic hydrocarbon derivatives.
2. The method according to claim 1, characterized in that Before obtaining N secondary mass spectra of the analyte, the method further comprises: The mass spectrum data of the analyte is deconvoluted and purified to obtain the N secondary mass spectra.
3. The method according to claim 1, characterized in that The machine learning model includes a normalization model and a random forest model; The feature recognition is performed on each of the N secondary mass spectra by the machine learning model to obtain a feature distribution evaluation result of M features corresponding to each secondary mass spectra, including: By using the normalization model, normalizing each of the N secondary mass spectra to obtain a normalized secondary mass spectrum corresponding to each secondary mass spectrum; The normalized secondary mass spectrum is subjected to feature recognition through the random forest model to obtain feature distribution evaluation results of M features corresponding to the normalized secondary mass spectrum.
4. The method according to claim 3, characterized in that Each of the N secondary mass spectra corresponds to a mass-to-charge ratio of a molecular ion peak; The normalization model is used to normalize each of the N secondary mass spectra to obtain a normalized secondary mass spectrum corresponding to each secondary mass spectrum, including: Obtaining the summary ion intensity of all fragments of the N secondary mass spectra according to the accumulated value of the ion intensity corresponding to each of the N secondary mass spectra; The summary mass-to-charge ratio related to the summary intensity of the ions and the mass-to-charge ratio of the fragments of each secondary mass spectrum are normalized to obtain a normalized secondary mass spectrum corresponding to each secondary mass spectrum.
5. The method according to claim 3 or 4, characterized in that: The M features include at least one of the following: regional intensity ratio, skewness, entropy, base-peak ratio; Wherein, the regional intensity ratio is used to characterize the proportion of the ion signal intensity of each secondary mass spectrogram to the total ion signal intensity of the N secondary mass spectrograms; The skewness is used to characterize the symmetry of the mass-to-charge ratio of each secondary mass spectrum in the mass-to-charge ratios of the N secondary mass spectrum fragments; The entropy is used to characterize the degree of disorder of the mass-to-charge ratio of each secondary mass spectrum in the mass-to-charge ratio of the N secondary mass spectrum fragments; The base peak ratio is used to characterize the relative degree of the base peak mass-to-charge ratio and the molecular ion peak mass-to-charge ratio in each secondary mass spectrum.
6. The method according to claim 5, characterized in that The M features are the regional intensity ratios, and the feature distribution results include a first feature distribution result corresponding to the regional intensity ratios; The method of performing feature recognition on the normalized secondary mass spectrum through the random forest model to obtain feature distribution evaluation results of M features corresponding to the normalized secondary mass spectrum includes: By means of the random forest model, the ion signal intensity of the normalized secondary mass spectrometer and the total ion signal intensity of the N normalized secondary mass spectrometers are identified; wherein the total ion signal intensity of the N normalized secondary mass spectrometers is the total ion signal intensity of the N secondary mass spectrometers; The ratio of the ion signal intensity of the normalized secondary mass spectrometer to the total ion signal intensity of the N normalized secondary mass spectrometers is determined as the first characteristic distribution result.
7. The method according to claim 5, characterized in that The M features are the skewness, the feature distribution result includes a second feature distribution result corresponding to the skewness, each of the N secondary mass spectra corresponds to a mass-to-charge ratio, and the mass-to-charge ratio corresponding to the normalized secondary mass spectrum is the mass-to-charge ratio corresponding to the secondary mass spectrum corresponding to the normalized secondary mass spectrum; The method of performing feature recognition on the normalized secondary mass spectrum by using the random forest model to obtain feature distribution evaluation results of M features corresponding to the normalized secondary mass spectrum includes: Calculate the difference between the mass-to-charge ratio corresponding to the normalized secondary mass spectrum and the average mass-to-charge ratio according to the mass-to-charge ratio corresponding to the normalized secondary mass spectrum and the average mass-to-charge ratio of N normalized secondary mass spectra; The ratio of the difference value to the product of the standard deviation of the mass-to-charge ratios corresponding to the N normalized secondary mass spectra and the first quantity of the normalized secondary mass spectra is determined as the second characteristic distribution result; wherein the first quantity takes a value of N-1.
8. The method according to claim 5, characterized in that The M features are the entropy, and the feature distribution result includes a third feature distribution result corresponding to the entropy; The method of performing feature recognition on the normalized secondary mass spectrum by using the random forest model to obtain feature distribution evaluation results of M features corresponding to the normalized secondary mass spectrum includes: Identify the relative ion signal intensities of P ion peaks in the normalized secondary mass spectrum, where P is an integer greater than 1; According to the relative ion signal intensity of each ion peak in the P ion peaks, the aggregate relative ion signal intensity of the P ion peaks is counted; The ion signal intensity after the summary relative ion signal intensity is normalized is determined as the third characteristic distribution result.
9. The method according to claim 1, characterized in that: The step of screening a first target secondary mass spectrum from the N secondary mass spectra comprises: According to the isotope difference value of the mass-to-charge ratio between the doubly charged fragments, a candidate secondary mass spectrum is screened from the N secondary mass spectra, wherein the candidate secondary mass spectrum is a doubly charged secondary mass spectrum; According to the first mass-to-charge ratio corresponding to the candidate secondary mass spectrum, calculating a second mass-to-charge ratio corresponding to the first mass-to-charge ratio, wherein the second mass-to-charge ratio is a mass-to-charge ratio corresponding to a single-charged secondary mass spectrum matching the candidate secondary mass spectrum; Based on the second mass-to-charge ratio, screening a single-charged secondary mass spectrum matching the candidate secondary mass spectrum from the N secondary mass spectra; The singly charged secondary mass spectrum and the candidate secondary mass spectrum are determined as the first target secondary mass spectrum.
10. The method according to claim 1, characterized in that The step of generating a molecular formula list of the analyte according to the feature distribution evaluation results of the M features corresponding to each secondary mass spectrum and the first target secondary mass spectrum comprises: According to the feature distribution evaluation results of the M features corresponding to each secondary mass spectra, a second target secondary mass spectra is selected from the N secondary mass spectra, and the feature distribution evaluation results of the M features corresponding to the second target secondary mass spectra meet the preset conditions; Molecular formula prediction is performed on the first target secondary mass spectrum and the second target secondary mass spectrum to obtain a molecular formula list of the analyte.
Citation Information
Cited By
Intelligent evaluation method for emission flux of organic matters in air
CN120877958A