Easily-mixed bacterium identification method, device, equipment and medium

Abstract

The invention discloses an easily-mixed bacterium identification method, device, equipment and medium, and relates to the technical field of microbiological identification, and the method comprises the following steps: obtaining a mass spectrum identification result sent by an original mass spectrometer, determining whether easily-mixed bacteria exist in the mass spectrum identification result, and if the easily-mixed bacteria exist in the mass spectrum identification result, determining whether the easily-mixed bacteria exist in the mass spectrum identification result. If yes, judging whether a micro-difference analysis model of which the type is the same as that of the miscible bacteria exists in a model library or not; if the micro-difference analysis model with the same type as the easily-mixed bacteria exists in the model library, calling the micro-difference analysis model to analyze and identify the easily-mixed bacteria so as to obtain a model analysis and identification result; and if the micro-difference analysis model with the same type as the easily-mixed bacteria does not exist in the model library, performing dynamic operation identification on the easily-mixed bacteria by adopting a dynamic matching algorithm, or performing dynamic operation identification on the easily-mixed bacteria by adopting the dynamic matching algorithm and calling the micro-difference analysis model to obtain a dynamic operation identification result, and the accuracy and efficiency of identifying similar microorganisms by the matrix-assisted laser desorption time-of-flight mass spectrometer are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of microbial identification, and in particular to a method, device, equipment and medium for identifying easily confusable bacteria. Background Art

[0002] Currently, MALDI-TOF MS (Matrix-Assisted Laser Desorption / Ionization TimeOf Flight Massspectrometry) is used for microbial identification. It mainly searches the database through fingerprints using a certain matching algorithm, and gives the identification results after similarity scoring. There are several types of widely used algorithms: (1) Expert spectrum identifier system based on strain-level database construction, which uses a weight matrix algorithm and ultimately scores based on the degree of weight similarity to give a similarity percentage; (2) Through the super spectrum comparison algorithm, the spectrum of the test bacteria is compared one by one with all reference spectra, and a reference identification result is given; (3) Unsupervised pattern matching algorithm, which defines the results of a certain score range as species-reliable, genus-reliable, and unreliable results based on big data calculations. However, whether using a weight matrix, a super map, or unsupervised pattern matching, in order to ensure that the instrument can still give relatively accurate results under the influence of various factors such as current industrial manufacturing, experimental conditions, and sample conditions, a scoring interval (similarity percentage) is used when judging the map results, and the results within a certain interval are specified as species reliable, a certain interval as genus reliable, and a certain interval as unreliable. Under this premise, a situation will inevitably arise, that is, when two or more microorganisms are closely related or the collected maps are similar (that is, the positions of high-intensity peaks in the maps are basically the same), the maps are evaluated through the current algorithm, which often leads to the final evaluation results that the scores of several spectrally similar bacteria are in the same confidence interval, making it difficult to accurately identify.

[0003] As can be seen from the above, how to improve the accuracy and efficiency of matrix-assisted laser desorption time-of-flight mass spectrometry in identifying similar microorganisms is a problem to be solved in this field. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a method, device, equipment and medium for identifying easily confusable bacteria, which can improve the accuracy and efficiency of matrix-assisted laser desorption time-of-flight mass spectrometry in identifying similar microorganisms. The specific scheme is as follows:

[0005] In a first aspect, the present application discloses a method for identifying easily confusable bacteria, comprising:

[0006] Obtaining the mass spectrometry identification result sent by the original mass spectrometer, determining whether there are easily confused bacteria in the mass spectrometry identification result, and if there are easily confused bacteria in the mass spectrometry identification result, determining whether there is a micro-difference analysis model of the same type as the easily confused bacteria in the preset model library;

[0007] If there is a micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, the micro-difference analysis model is retrieved to analyze and identify the easily confused bacteria to obtain a model analysis and identification result;

[0008] If there is no micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, a dynamic matching algorithm is used to perform dynamic calculation identification on the easily confused bacteria, or a dynamic matching algorithm is used and the micro-difference analysis model is called to perform dynamic calculation identification on the easily confused bacteria to obtain a dynamic calculation identification result.

[0009] Optionally, before determining whether there is a micro-difference analysis model of the same type as the easily mixed bacteria in the preset model library, the method further includes:

[0010] The micro-difference analysis method was used to construct each micro-difference analysis model;

[0011] Each of the micro-difference analysis models is saved in the model library.

[0012] Optionally, the micro-difference analysis method is used to construct each micro-difference analysis model, including:

[0013] Acquire historical data, adopt characteristic value analysis method and use preset characteristic value analysis system to analyze the historical data to obtain fingerprints of various strains, process and perform big data analysis on the fingerprints of various strains to construct micro-difference analysis models;

[0014] Alternatively, a graph trend analysis method and an artificial intelligence learning method are used to autonomously learn the bacterial species types in the historical data to obtain the graph trend differences between different groups, a reference graph trend differentiation model is constructed based on the graph trend differences, and the reference graph trend differentiation model is verified and corrected to obtain each of the micro-difference analysis models.

[0015] Optionally, the micro-difference analysis model includes a chain analysis model; and the construction process of the chain analysis model includes:

[0016] Obtaining a protein fingerprint of a known strain that is easily confused with a mass spectrometer, and performing noise processing on the protein fingerprint to obtain the processed protein fingerprint;

[0017] The peak values ​​of different species in the processed protein fingerprint map are multiplexed and compared to obtain the special unique peak values ​​corresponding to different species, and the special unique peak values ​​are classified using a preset binary chain analysis model to construct a chain analysis model, and the chain analysis model is trained to obtain a target chain analysis model.

[0018] Optionally, the peak multiple analysis and comparison of different species in the processed protein fingerprint to obtain each special unique peak corresponding to different species includes:

[0019] The processed protein fingerprint map is subjected to multiple analyses using a preset typing software, and multiple peak comparisons are performed on different species in the processed protein fingerprint map to obtain the specific unique peaks corresponding to different species.

[0020] Optionally, the training of the chain analysis model to obtain a target chain analysis model includes:

[0021] Acquire historical species protein fingerprint peaks, and use the historical species protein fingerprint peaks to train the chain analysis model to obtain training results;

[0022] The parameters in the chain analysis model are modified and adjusted based on the training results to obtain a target chain analysis model.

[0023] Optionally, the process of using a dynamic matching algorithm to perform dynamic calculation and identification on the easily mixed bacteria is:

[0024] Setting an initial tolerance value for mass spectrum matching, and matching the mass spectrum of easily mixed bacteria to be matched with the mass spectrum database to obtain an initial matching result, an initial peak matching situation, and a database spectrum corresponding to the initial tolerance value;

[0025] Using the initial matching result and the initial peak matching situation, judging whether the initial tolerance value satisfies a preset convergence condition, and if the initial tolerance value satisfies the preset convergence condition, taking the initial matching result as the dynamic calculation identification result of the easily confusable bacteria;

[0026] If the initial tolerance value does not meet the preset convergence condition, the initial matching result and the initial peak matching condition are calculated to obtain an optimized tolerance value;

[0027] The database spectrum is matched based on the optimized tolerance value to obtain the current matching result and the current peak matching situation, and the process of judging the convergence condition is repeatedly executed until the optimized tolerance value satisfies the convergence condition, and the current matching result is used as the dynamic calculation identification result of the easily confused bacteria.

[0028] Optionally, the setting of an initial tolerance value for mass spectrum matching, and matching the mass spectrum of easily mixed bacteria to be matched with a mass spectrum database to obtain an initial matching result, an initial peak matching situation, and a database spectrum corresponding to the initial tolerance value, includes:

[0029] Adaptively set the initial tolerance value for mass spectrum matching;

[0030] The mass spectrum to be matched of the easily mixed bacteria is matched with all spectra in the mass spectrum database to obtain an initial matching result, an initial peak matching situation and a database spectrum corresponding to the initial tolerance value; wherein the initial peak matching situation includes the error values ​​of all matching peaks in the mass spectrum database and the mass spectrum peaks to be matched of the mass spectrum to be matched.

[0031] Optionally, the using the initial matching result and the initial peak matching situation to judge whether the initial tolerance value satisfies a preset convergence condition includes:

[0032] A termination equation is set, the initial matching result and the initial peak matching condition are input into the termination equation, and it is determined whether the initial tolerance value meets the preset convergence condition; the termination equation is:

[0033] ;

[0034] in, is the initial matching result, This is the initial peak matching situation.

[0035] Optionally, the determining whether the initial tolerance value satisfies a preset convergence condition, if the initial tolerance value satisfies the preset convergence condition, includes:

[0036] Determine whether all mass spectrum peak error values ​​in the initial matching result are not greater than the initial tolerance value;

[0037] If all mass spectrum peak error values ​​in the initial matching result are not greater than the initial tolerance value, the initial peak matching situation is uniquely identified. If the uniqueness identification passes, it is determined that the initial tolerance value meets the preset convergence condition.

[0038] Optionally, before calculating the initial matching result and the initial peak matching situation, the method further includes:

[0039] Set the intermediate variable matching results and intermediate variable peak matching conditions;

[0040] The values ​​of the initial matching result and the initial peak matching situation are assigned to the intermediate variable matching result and the intermediate variable peak matching situation.

[0041] Optionally, the calculating the initial matching result and the initial peak matching condition to obtain an optimized tolerance value includes:

[0042] A tolerance adjustment equation is set, and the intermediate variable matching result and the intermediate variable peak matching situation are input into the tolerance adjustment equation for calculation to obtain an optimized tolerance value; the tolerance adjustment equation is:

[0043] ;

[0044] in, To optimize the tolerance value, is the intermediate variable matching result, This is the peak matching situation of the intermediate variable.

[0045] Optionally, the dynamic matching algorithm is used to perform dynamic calculation and identification on the easily mixed bacteria, including:

[0046] Determine whether there is a target dynamic operation start rule corresponding to each dynamic operation start rule in the dynamic operation start rule base;

[0047] If there is a target dynamic operation start rule corresponding to each of the dynamic operation start rules in the dynamic operation start rule library, the target dynamic operation start rule is started, and a dynamic matching algorithm is used to perform dynamic operation identification on easily confusable bacteria.

[0048] Optionally, before determining whether there is a target dynamic operation activation rule corresponding to each dynamic operation activation rule in the dynamic operation activation rule base, the method further includes:

[0049] Build dynamic operation startup rules based on business needs and in accordance with the preset rule building method;

[0050] Each of the dynamic operation activation rules is saved in the preset dynamic operation activation rule library.

[0051] Optionally, the adopting of a dynamic matching algorithm and calling the micro-difference analysis model to perform dynamic computational identification on the easily confusable bacteria to obtain a dynamic computational identification result includes:

[0052] Setting an initial tolerance value for mass spectrum matching, performing a first round of dynamic calculation, matching the mass spectrum of easily mixed bacteria to be matched with the mass spectrum database, so as to obtain an initial matching result, an initial peak matching situation, and a database spectrum corresponding to the initial tolerance value;

[0053] Using the initial matching result and the initial peak matching situation, it is judged whether the initial tolerance value satisfies the preset convergence condition. If the initial tolerance value satisfies the preset convergence condition, the dynamic operation ends, and the initial matching result is used as the dynamic operation identification result of the easily confusable bacteria;

[0054] If the initial tolerance value does not meet the preset convergence condition, that is, the first round of dynamic calculation cannot obtain an identification result, then determine whether there is a micro-difference analysis model of the same type as the current matching result in the preset model library; if there is a micro-difference analysis model of the same type as the current matching result in the preset model library, then call the micro-difference analysis model to analyze and identify the current matching result to obtain a model analysis and identification result;

[0055] If there is no micro-difference analysis model of the same type as the current matching result in the preset model library, the initial matching result and the initial peak matching situation are calculated to obtain an optimized tolerance value;

[0056] A new round of dynamic calculation and model matching process is repeated based on the optimized tolerance value until an identification result is obtained.

[0057] In a second aspect, the present application discloses a device for identifying easily confusable bacteria, comprising:

[0058] A judgment module is used to obtain the mass spectrometry identification result sent by the original mass spectrometer, and judge whether there are easily mixed bacteria in the mass spectrometry identification result. If there are easily mixed bacteria in the mass spectrometry identification result, it is judged whether there is a micro-difference analysis model of the same type as the easily mixed bacteria in the preset model library;

[0059] A model analysis module is used for, if there is a micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, calling the micro-difference analysis model to analyze and identify the easily confused bacteria, so as to obtain a model analysis and identification result;

[0060] The dynamic operation and identification module is used to use a dynamic matching algorithm to perform dynamic operation and identification on the easily confused bacteria if there is no micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, or to use a dynamic matching algorithm and call the micro-difference analysis model to perform dynamic operation and identification on the easily confused bacteria to obtain a dynamic operation and identification result.

[0061] In a third aspect, the present application discloses an electronic device, comprising:

[0062] Memory, used to store computer programs;

[0063] The processor is used to execute the computer program to implement the aforementioned method for identifying easily confusable bacteria.

[0064] In a fourth aspect, the present application discloses a computer storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the steps of the above-disclosed method for identifying easily confusable bacteria are implemented.

[0065] It can be seen that the present application provides a method for identifying easily confused bacteria, including obtaining a mass spectrometer identification result sent by an original mass spectrometer, determining whether easily confused bacteria exist in the mass spectrometer identification result, and if easily confused bacteria exist in the mass spectrometer identification result, determining whether there is a micro-difference analysis model of the same type as the easily confused bacteria in a preset model library; if there is a micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, calling the micro-difference analysis model to analyze and identify the easily confused bacteria to obtain a model analysis identification result; if there is no micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, using a dynamic matching algorithm to perform dynamic calculation identification on the easily confused bacteria, or using a dynamic matching algorithm and calling the micro-difference analysis model to perform dynamic calculation identification on the easily confused bacteria to obtain a dynamic calculation identification result. The present application first determines whether there are easily confused bacteria in the mass spectrometry identification results sent by the original mass spectrometer. If there are easily confused bacteria, it is determined whether there is a micro-difference analysis model of the same type as the easily confused bacteria in the preset model library. If there is a micro-difference analysis model of the same type as the easily confused bacteria, the differential difference analysis model is retrieved to analyze and identify the easily confused bacteria to obtain a model analysis and identification result; if there is no micro-difference analysis model of the same type as the easily confused bacteria, a dynamic matching algorithm is used to perform dynamic calculation identification on the easily confused bacteria, or a dynamic matching algorithm is used and the micro-difference analysis model is retrieved to perform dynamic calculation identification on the easily confused bacteria to obtain a dynamic calculation identification result. The present application realizes a multi-faceted process of identifying easily confused bacteria through a differential difference analysis model and a dynamic matching algorithm, thereby improving the accuracy and efficiency of matrix-assisted laser desorption time-of-flight mass spectrometer in identifying similar microorganisms. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0067] Figure 1 This is a flow chart of a method for identifying easily mixed bacteria disclosed in this application;

[0068] Figure 2 This is a flow chart of a specific implementation method of the first multiple analysis and comparison disclosed in this application;

[0069] Figure 3 This is a flow chart of a specific implementation method of the second multiple analysis and comparison disclosed in this application;

[0070] Figure 4 This is a schematic diagram of the structure of an easily mixed bacteria identification device disclosed in this application;

[0071] Figure 5 A structural diagram of an electronic device provided for this application. DETAILED DESCRIPTION

[0072] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0073] At present, MALDI-TOF MS is used for microbial identification, mainly through fingerprints, searching the database with a certain matching algorithm, and giving the identification results after similarity scoring. There are several types of widely used algorithms: (1) Expert spectrum identifier system based on species-level database construction, using weight matrix algorithm, and finally scoring according to the degree of weight similarity to give a similarity percentage; (2) Super spectrum comparison algorithm, using the test bacteria spectrum and all reference spectra to compare one by one, and give a reference identification result; (3) Unsupervised pattern matching algorithm, based on big data calculation, defines the results of a certain score interval as species reliable, genus reliable and unreliable results. However, whether using weight matrix, super spectrum, or unsupervised pattern matching, in order to ensure that the instrument can still give relatively accurate results under the influence of various factors such as current industrial manufacturing, experimental conditions, sample conditions, etc., a scoring interval (similarity percentage) is used when judging the spectrum results, and it is stipulated that the results within a certain interval are species reliable, a certain interval is genus reliable, and a certain interval is unreliable. Under this premise, a situation will inevitably occur, that is, when two or more microorganisms are closely related or the collected spectra are similar (that is, the high-intensity peaks in the spectra are basically the same), the spectra are evaluated by the current algorithm, which often leads to the final evaluation results that several bacteria scores are in the same confidence interval, making it difficult to accurately identify. As can be seen from the above, how to improve the accuracy and efficiency of matrix-assisted laser desorption time-of-flight mass spectrometry in identifying similar microorganisms is a problem to be solved in this field.

[0074] See also Figure 1 As shown, the embodiment of the present invention discloses a method for identifying easily confusable bacteria, which may specifically include:

[0075] Step S11: Obtain the mass spectrometry identification result sent by the original mass spectrometer, and determine whether there are easily confused bacteria in the mass spectrometry identification result. If there are easily confused bacteria in the mass spectrometry identification result, determine whether there is a micro-difference analysis model of the same type as the easily confused bacteria in the preset model library.

[0076] The easily confused bacteria in this application are bacteria with close kinship, multiple bacteria belonging to the same complex group or two or more bacteria with similar other protein fingerprint spectra, and their protein fingerprint feature spectra are highly similar, that is, there are at least two bacteria with identification score values ​​in the same confidence interval. In this case, different types of bacteria in the same confidence interval (i.e., different species, or different genera) are called mass spectrometry easily confused bacteria or mass spectrometry difficult to distinguish bacteria, referred to as easily confused bacteria. The confidence interval is generally divided into multiple sections according to the identification score value, such as: species level credible, genus level credible and untrustworthy.

[0077] In this embodiment, before determining whether there is a micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, it also includes: constructing each micro-difference analysis model using a micro-difference analysis method; and saving each micro-difference analysis model to the model library. Among them, the micro-difference analysis method is used to construct each micro-difference analysis model, including: obtaining historical data, using a characteristic value analysis method and using a preset characteristic value analysis system to analyze the historical data to obtain the distribution and intensity of the micro-difference peaks of each bacterial species fingerprint, and establishing a primary micro-difference model based on the micro-difference peaks, using big data analysis combined with a rolling adjustment model parameter method to correct the accuracy and precision of the micro-difference model to determine each of the micro-difference analysis models; or, using a spectrum trend analysis method and an artificial intelligence learning method to autonomously learn the bacterial species types in the historical data to obtain the spectrum trend differences between different groups, constructing a reference spectrum trend differentiation model based on the spectrum trend differences, and verifying and correcting the reference spectrum trend differentiation model to obtain each of the micro-difference analysis models.

[0078] There are two methods for micro-difference analysis: eigenvalue analysis and graph trend analysis. The implementation process is as follows: (1) Eigenvalue analysis: Use the self-developed eigenvalue analysis system to analyze the data, compare the fingerprint spectra of easily confused bacteria, shield the same protein fingerprint peaks, shield the unstable protein fingerprint peaks, and select the protein fingerprint peaks with high abundance, strong stability, high peak intensity, and large relative abundance of peaks from the remaining protein fingerprint peaks as reference difference points. Then perform big data analysis and verification, during which the number of reference difference points and the model parameters that affect accuracy and precision are adjusted in a rolling manner to determine the optimal difference point for distinguishing and identifying, and determine the detection tolerance, relative abundance, relative abundance tolerance and other parameters of the characteristic peak of the optimal difference point. Based on the above big data analysis results, a differentiation model for easily confused bacteria is established and incorporated into the model library.

[0079] (2) Spectrum trend analysis method: The spectra are classified into different groups according to the type of bacteria. Through the artificial intelligence self-learning algorithm, autonomous learning is performed among multiple groups to distinguish the differences in the spectrum trends between different groups. Based on these trend differences, a reference spectrum trend differentiation model for easily confused bacteria is established through self-encoding. After big data verification and continuous model correction, the optimal spectrum trend analysis model is finally determined. The model is then included in the model library.

[0080] In this embodiment, if there are easily confused bacteria in the mass spectrometry identification results, it is determined whether there is a micro-difference analysis model of the same type as the easily confused bacteria in the preset model library; the micro-difference analysis model at least includes a chain analysis model; the construction process of the chain analysis model includes: obtaining a protein fingerprint of a known strain that is easily confused with a mass spectrometry bacteria, and performing noise reduction processing on the protein fingerprint to obtain the processed protein fingerprint; performing peak multiple alignment analysis on different species in the processed protein fingerprint to obtain each special unique peak corresponding to different species, using a preset binary chain analysis model to classify each of the special unique peaks, and constructing a chain analysis model for all strains participating in the analysis according to the binary classification rule of gradually reducing the number of populations, and using the protein fingerprint of the training set strain to train the constructed chain analysis model to obtain a target chain analysis model.

[0081] Specifically, microbial protein fingerprints are collected, and protein fingerprints of known strains that are easy to be easily mixed with bacteria by mass spectrometry and meet preset conditions for easy mixing with bacteria are screened from all the microbial protein fingerprints; the protein fingerprints are processed for noise using preset typing software to obtain the processed protein fingerprints.

[0082] The specific noise processing process is: inputting the protein fingerprint spectrum into a preset typing software, and performing noise processing on the protein fingerprint spectrum based on the wavelet transform method; the noise processing includes exponential smoothing elimination processing.

[0083] The present application removes noise in the spectrum through wavelet transform, eliminates burr peaks caused by system errors through exponential smoothing, and reduces unstable interference factors included in the model group.

[0084] The peak multiple analysis and comparison process is as follows: using the preset typing software to perform multiple analysis on all processed protein fingerprints, and performing multiple comparison of peaks of different species in the processed protein fingerprints to obtain the special unique peaks corresponding to different species;

[0085] The process of training the chain analysis model is as follows: obtaining historical species protein fingerprint peaks, using the historical species protein fingerprint peaks to train the chain analysis model to obtain training results; based on the training results, modifying and adjusting the parameters in the chain analysis model to improve the model accuracy and specificity to obtain the target chain analysis model.

[0086] The specific process of constructing the chain analysis model in the present application is: based on each special unique peak, a single binary analysis model is established in the typing software. When the binary analysis model is established, the peak number, peak matching tolerance, peak abundance, peak abundance tolerance and other parameters should be preliminarily set according to the analysis results of the medium-sized data volume of the protein fingerprint map. Then, the established binary analysis model is cross-correlated to form a chain analysis structure, and a chain analysis model is constructed for the joint analysis of multiple easily confused bacteria.

[0087] The specific implementation methods of multiple analysis and comparison in this application include the following two situations:

[0088] (1) The specific implementation process of the first type of multiple analysis and comparison is as follows Figure 2 As shown, an identification result is set as a genus level identification, which includes a common characteristic peak. First, a species with a special unique peak and the peak of the genus level identification are taken to form the first binary judgment. There are two results. One is that it contains a special peak and can be judged as the first species, and the other is that it does not contain a characteristic peak and can only be judged as the name of the genus. At this time, the second species with a characteristic peak is taken to form a second binary judgment with the genus identification, and so on until the analysis tree reaches indivisibility. At this time, the distinction of all easily confused identification bacteria can be completed;

[0089] (2) The specific implementation process of the second multiple analysis and comparison is as follows Figure 3 As shown, firstly, several indistinguishable bacteria are subjected to group peak analysis, and the bacteria are divided into two groups according to the distribution of characteristic peaks, and the difference peak combination between the two groups is used as the first binary branch judgment; then, peak analysis is performed again in the two groups, and the bacteria in the groups are divided into two small bacteria through specific peak combinations, forming the second binary branch judgment; and so on until they are finally indistinguishable.

[0090] In addition, the present application uses known historical species protein fingerprint peaks for model training, and modifies the peak judgment parameters in the model according to the training results until the model accuracy is stable, thereby obtaining the target chain analysis model.

[0091] In this embodiment, the target chain analysis model is modularly stored, and the model startup rules are set to obtain the protein fingerprint spectrum of the easily confused bacteria to be identified, and determine whether the protein fingerprint spectrum meets the model startup rules. If the protein fingerprint spectrum meets the model startup rules, the target chain analysis model is started, and the protein fingerprint spectrum is input into the target chain analysis model for identification, so as to output the identification result of the easily confused bacteria to be identified.

[0092] In other words, the target chain analysis model is stored modularly, and the model activation rules are defined. For example, when mixed identifications at the species level appear in the identification results, and the mixed identifications happen to be species that can be distinguished by the model, the chain model is retrieved from the stored model library and the results are accurately interpreted.

[0093] The advantages of the micro-difference analysis model are: (1) it can be distinguished by establishing a model to reduce the dependence on the map; (2) it can accurately identify the easily confused identification strains included in the model; (3) it can be applied not only to the accurate distinction of two-to-two mixed identification bacteria, but also to the accurate identification of multiple mixed identification bacteria. This application uses the difference points between strains and groups to connect the easily confused identification species through the difference points in series, and accurately identifies each strain in the easily confused identification strain group.

[0094] Step S12: If there is a micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, the micro-difference analysis model is retrieved to analyze and identify the easily confused bacteria to obtain a model analysis and identification result.

[0095] In this embodiment, the easily confused bacteria are matched with the model library, the micro-difference analysis model is called to analyze and identify the easily confused bacteria, and the model analysis and identification result of the single species name at the species level is output: when the easily confused bacteria exist in the mass spectrometry identification result obtained by the mass spectrometer, and the species of the easily confused bacteria is the same as a certain micro-difference analysis model in the model library, the corresponding micro-difference analysis model in the model library is called to perform analysis and identification, that is, further fine-analyze the easily confused bacteria in the original report, and obtain the model analysis and identification result of the single species name at the species level. That is, let only one species have a score greater than the minimum tolerance value of species-level identification.

[0096] For example, for the easily confused bacteria found in the MALDI-TOF MS identification process, a small model is constructed as a small unit. In this model unit, because the amount of data is relatively small, it can be distinguished one by one through the micro-difference analysis method based on the subtle differences in the protein fingerprints of a limited number of specific bacterial species; for example, model 1 is constructed for easily confused species A and B, model 2 is constructed for easily confused species A and C, model 3 is constructed for easily confused species A, D and E, etc. These models can be used alone or in combination. For example, the aforementioned model unit 1 is combined with model unit 2 to distinguish species A, B, C, etc. Finally, all the combinations formed by these unitizations or unit links are included in a cluster to form a model library; the activation rule of the model library is: when the species-level identification results in the mass spectrometry identification results include and only include species A and B, the identification of model 1 in the model library is triggered, and model 1 is called to accurately distinguish between species A and B. If the model determines that the result of species A is the final result, then in the final report, the result of species A is retained at the species-level identification, that is, the score of species A is above the minimum tolerance value for species-level identification, and the result of species B is reduced to the genus-level identification, that is, the score of strain B drops below the minimum tolerance value for species-level identification; if the species-level identification results in the original report results only include species A, B, and C, then the chain model composed of model 1 and model 2 in the model library is triggered, and the chain model is called to further analyze species A, B, and C, and generate model analysis and identification results.

[0097] Step S13: If there is no micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, a dynamic matching algorithm is used to perform dynamic calculation identification on the easily confused bacteria, or a dynamic matching algorithm is used and the micro-difference analysis model is called to perform dynamic calculation identification on the easily confused bacteria to obtain a dynamic calculation identification result.

[0098] In this embodiment, the process of using the dynamic matching algorithm to perform dynamic calculation and identification on the easily mixed bacteria is as follows:

[0099] Setting an initial tolerance value for mass spectrum matching, and matching the mass spectrum of easily mixed bacteria to be matched with the mass spectrum database to obtain an initial matching result, an initial peak matching situation, and a database spectrum corresponding to the initial tolerance value;

[0100] Using the initial matching result and the initial peak matching situation, judging whether the initial tolerance value satisfies a preset convergence condition, and if the initial tolerance value satisfies the preset convergence condition, taking the initial matching result as the dynamic calculation identification result of the easily confusable bacteria;

[0101] If the initial tolerance value does not meet the preset convergence condition, the initial matching result and the initial peak matching condition are calculated to obtain an optimized tolerance value;

[0102] The database spectrum is matched based on the optimized tolerance value to obtain the current matching result and the current peak matching situation, and the process of judging the convergence condition is repeatedly executed until the optimized tolerance value satisfies the convergence condition, and the current matching result is used as the dynamic calculation identification result of the easily confused bacteria.

[0103] Among them, the process of setting the initial tolerance value for mass spectrum matching and matching the mass spectrum to be matched of easily confused bacteria with the mass spectrum database is: adaptively setting the initial tolerance value for mass spectrum matching; matching the mass spectrum to be matched of easily confused bacteria with all spectra in the mass spectrum database to obtain the initial matching results, initial peak matching conditions and database spectra corresponding to the initial tolerance value; wherein the initial peak matching conditions include the error values ​​between all matching peaks in the mass spectrum database and the mass spectrum peaks to be matched of the mass spectrum to be matched.

[0104] In the process of mass spectrometry matching and identification of microorganisms, this application introduces a tolerance value t when matching mass spectrum peaks in order to correct the mass number error generated during instrument detection. Tolerance value: refers to the selection range when matching peaks. The larger the tolerance, the larger the range selected when matching peaks. The purpose of the tolerance is that under certain authorizations, small changes in the number of peaks will not interfere with the accuracy of the overall matching results. Usually t is an appropriate constant to ensure that the mass spectrum peak is not misjudged due to t being too large, and the mass spectrum peak is not missed due to t being too small. However, due to certain inter-instrument differences on the instrument, and different differences in fingerprint spectra of different microorganisms or different strains of microorganisms, the pre-set tolerance value t may have a certain impact on the correctness of the actual matching results. This application proposes a quantitative method of adaptive t, so that the tolerance value t can be automatically optimized during the matching process, and finally a suitable value is used for matching judgment.

[0105] In this embodiment, the initial tolerance value for mass spectrum matching is adaptively set; the mass spectrum to be matched of easily confused bacteria is matched with all spectra in the mass spectrum database to obtain an initial matching result, an initial peak matching situation, and a database spectrum corresponding to the initial tolerance value; wherein the initial peak matching situation includes the error values ​​between all matching peaks in the mass spectrum database and the mass spectrum peak to be matched of the mass spectrum to be matched.

[0106] Specifically, use a larger tolerance value As the initial value, get the initial tolerance value , the mass spectrum to be matched Spectra in the mass spectrometry database Perform matching to obtain the initial sorted matching results ,and Corresponding database map , and the initial peak matching (in, Including the error values ​​between all mass spectrum database matching peaks and the mass spectrum peaks to be matched).

[0107] The process of judging whether the initial tolerance value satisfies the preset convergence condition is as follows: setting a termination equation, inputting the initial matching result and the initial peak matching condition into the termination equation; the termination equation is:

[0108] ;

[0109] in, is the initial matching result, This is the initial peak matching situation.

[0110] The convergence condition judgment process is as follows: determine whether all mass spectrum peak error values ​​in the initial matching result are not greater than the initial tolerance value; if all mass spectrum peak error values ​​in the initial matching result are not greater than the initial tolerance value, then uniquely identify the initial peak matching situation, that is, there is only one single bacterial species name in the matching result that is above the minimum tolerance value for species level identification. If the uniqueness identification is passed, it is determined that the initial tolerance value meets the preset convergence condition.

[0111] Specifically, a termination equation is set to determine the initial tolerance value Whether it has converged (i.e., judging is the best tolerance), and use this termination equation to Matching with the initial peak Determine the initial tolerance Whether the convergence condition has been reached. Matching with the initial peak Substitute into the termination equation and judge the convergence condition. If the tolerance is When the variable All mass spectrometry peak error values ​​are less than or equal to In this case, the matching result If the uniqueness is passed, it means that the tolerance has converged, that is, the initial tolerance value meets the preset convergence condition. Assigned to the tolerance variable after convergence ,Right now ,Will As the result of dynamic calculation identification of easily mixed bacteria , .

[0112] Then, it is determined whether all mass spectrum peak error values ​​in the initial matching result are not greater than the initial tolerance value; if all mass spectrum peak error values ​​in the initial matching result are not greater than the initial tolerance value, the initial peak matching situation is uniquely identified, that is, there is only one single bacterial species name in the matching result that is above the minimum tolerance value for species level identification. If the uniqueness identification is passed, it is determined that the initial tolerance value meets the preset convergence condition.

[0113] If the initial tolerance value does not meet the preset convergence condition, the intermediate variable matching result and the intermediate variable peak matching situation are set; the values ​​of the initial matching result and the initial peak matching situation are assigned to the intermediate variable matching result and the intermediate variable peak matching situation, and the tolerance adjustment equation is set, and the intermediate variable matching result and the intermediate variable peak matching situation are input into the tolerance adjustment equation for calculation to obtain the optimized tolerance value; the tolerance adjustment equation is:

[0114] ;

[0115] in, To optimize the tolerance value, is the intermediate variable matching result, This is the peak matching situation of the intermediate variable.

[0116] Specifically, if the initial tolerance value does not meet the preset convergence condition, then set the intermediate variable, including the intermediate variable matching result Matching with intermediate variable peaks , receive the initial match to obtain the initial match result Matching with the initial peak ,make , , then set the tolerance adjustment equation, the intermediate variable matches the result Matching with intermediate variable peaks Substitute into the tolerance adjustment equation to get the optimized tolerance value .

[0117] The database spectrum is matched based on the optimized tolerance value to obtain the current matching result and the current peak matching situation. Database Graph Match again to get the current matching result ,and Corresponding current database map Matching with current peak , again using the termination equation, Substitute into the termination equation for calculation and determine Whether it has converged. If it has converged, the corresponding current matching result will be As the result of dynamic calculation identification of easily mixed bacteria.

[0118] In this embodiment, an initial tolerance value for mass spectrum matching is set, and the mass spectrum to be matched of the easily confused bacteria is matched with the mass spectrum database to obtain an initial matching result, an initial peak matching situation and a database spectrum corresponding to the initial tolerance value; using the initial matching result and the initial peak matching situation, it is judged whether the initial tolerance value meets the preset convergence condition; if the initial tolerance value meets the preset convergence condition, the initial matching result is used as the dynamic operation identification result of the easily confused bacteria; if the initial tolerance value does not meet the preset convergence condition, the initial matching result and the initial peak matching situation are calculated to obtain an optimized tolerance value; based on the optimized tolerance value, the database spectrum is matched to obtain a current matching result and a current peak matching situation, and the process of judging the convergence condition is repeated until the optimized tolerance value meets the convergence condition, and the current matching result is used as the dynamic operation identification result of the easily confused bacteria. The present application uses the set initial tolerance value to match the mass spectrum to be matched with the mass spectrum database, and determines whether the initial tolerance value meets the preset convergence condition. If it does, the initial matching result is used as the dynamic operation identification result of the easily mixed bacteria. If it does not, the optimized tolerance value is calculated, and the matching is performed again until the optimized tolerance value meets the convergence condition, and the current matching result is used as the dynamic operation identification result of the easily mixed bacteria. The present application can realize the dynamic operation identification of easily mixed bacteria through the dynamic matching between the mass spectrum to be matched and the mass spectrum database and the convergence judgment of the tolerance value without adding an additional database, reduce the workload of the dynamic operation identification of the mixed bacteria mass spectrum, and is applicable to the identification problems of different easily mixed identification bacteria, reduce the complexity of the dynamic operation identification of easily mixed bacteria, and improve the efficiency and accuracy of the dynamic operation identification of easily mixed bacteria.

[0119] In this embodiment, if there is no micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, it is determined whether there is a target dynamic operation start rule corresponding to each dynamic operation start rule in the dynamic operation start rule library; if there is a target dynamic operation start rule corresponding to each dynamic operation start rule in the dynamic operation start rule library, the target dynamic operation start rule is started, and a dynamic matching algorithm is used to perform dynamic operation identification on the easily confused bacteria;

[0120] Among them, before determining whether there is a target dynamic operation startup rule corresponding to each dynamic operation startup rule in the dynamic operation startup rule library, it also includes: constructing each dynamic operation startup rule based on business needs and according to a preset rule construction method; and saving each of the dynamic operation startup rules to the preset dynamic operation startup rule library.

[0121] Therefore, the prerequisite for the process of dynamic operation identification of easily confused bacteria is: first, determine whether there is a target dynamic operation startup rule corresponding to each of the dynamic operation startup rules in the dynamic operation startup rule base locally; if there is a target dynamic operation startup rule corresponding to each of the dynamic operation startup rules in the dynamic operation startup rule base locally, then start the target dynamic operation startup rule and execute the process of dynamic operation identification of easily confused bacteria proposed in this application.

[0122] In a specific embodiment of the present invention, the process of using a dynamic matching algorithm and calling the micro-difference analysis model to perform dynamic operation and identification on the easily confused bacteria is as follows: setting an initial tolerance value for mass spectrum matching, performing a first round of dynamic operation, matching the mass spectrum to be matched of the easily confused bacteria with the mass spectrum database to obtain an initial matching result, an initial peak matching situation, and a database spectrum corresponding to the initial tolerance value; using the initial matching result and the initial peak matching situation, judging whether the initial tolerance value meets a preset convergence condition; if the initial tolerance value meets the preset convergence condition, the dynamic operation ends, and the initial matching result is used as the dynamic operation and identification result of the easily confused bacteria; if the initial tolerance value does not meet the preset convergence condition, the dynamic operation ends, and the initial matching result is used as the dynamic operation and identification result of the easily confused bacteria; Assuming a convergence condition, that is, the first round of dynamic operation cannot obtain an identification result, it is determined whether there is a micro-difference analysis model of the same type as the current matching result in the preset model library; if there is a micro-difference analysis model of the same type as the current matching result in the preset model library, the micro-difference analysis model is called to analyze and identify the current matching result to obtain a model analysis and identification result; if there is no micro-difference analysis model of the same type as the current matching result in the preset model library, the initial matching result and the initial peak matching situation are calculated to obtain an optimized tolerance value; based on the optimized tolerance value, a new round of dynamic operation and model matching process is repeated until an identification result is obtained.

[0123] Based on the MALDI-TOF MS identification results, this application combines the dynamic matching algorithm with the micro-difference analysis model in the model library, eliminates erroneous matching results through multi-dimensional calculations, and ultimately improves the accuracy of MALDI-TOF MS identification of microorganisms. Specifically, it can be divided into the following five situations:

[0124] (1) Directly use the micro-difference analysis model to analyze and identify easily confused bacteria. That is, when easily confused bacteria exist in the mass spectrometry identification results, and there is a micro-difference analysis model of the same type as the easily confused bacteria in the model library, directly use the micro-difference analysis model to analyze and identify the easily confused bacteria, call the corresponding micro-difference analysis model in the model library, further analyze the easily confused bacteria in the mass spectrometry identification results, and give an accurate species-level model analysis and identification result, that is, a single species name. (There is a model data of the corresponding easily confused bacteria in the model library. In this case, the identification accuracy of the confused bacteria is relatively high).

[0125] (2) Micro-difference analysis model + dynamic operation identification. That is, when there are easily confused bacteria in the mass spectrometry identification results, and there is a micro-difference analysis model of the same type as the easily confused bacteria in the model library, the operation method is the same as (1); when there are easily confused bacteria in the mass spectrometry identification results, and there is no micro-difference analysis model of the same type as the easily confused bacteria in the model library, the dynamic operation identification is started, and a round of dynamic analysis is performed on the easily confused bacteria. If the incorrect easily confused bacteria in the original matching results are removed after the dynamic operation (that is, after the first round of dynamic operation, only one species has a score greater than the minimum tolerance value of species-level identification), the final result is directly reported; if there are still at least two species in the species-level identification results after a round of dynamic operation, but at this time, the type of easily confused bacteria in the species-level identification results happens to be the same as a certain micro-difference analysis model in the model library, then the micro-difference analysis model analysis and identification is started, and the corresponding micro-difference analysis model is called for further analysis. The execution process is the same as in (1) and will not be repeated here. If at least two bacterial species still exist in the species-level identification results after one round of dynamic operation, and there is no matching micro-difference analysis model at this time, a second round of dynamic operation is performed on the easily confused bacteria, and then model matching and model analysis are attempted; the above actions are repeated many times until a corresponding micro-difference analysis model is successfully matched, or after multiple rounds of dynamic operation, the species level retains the only identified bacterial species, that is, there is only one bacterial species with a score higher than the minimum tolerance value of species-level identification, and the final dynamic operation identification result is output at this time; if multiple rounds of dynamic operation cannot match the micro-difference analysis model startup rules in the model library, and after multiple rounds of dynamic operation, there is no bacterial species in the species-level identification, that is, all bacterial species scores are less than the minimum tolerance value of species-level identification. At this time, the original matching result is directly returned, and prompts and instructions are given for the easily confused bacterial species in the original results.

[0126] (3) Micro-difference analysis model + judgment on compliance with dynamic operation start-up rules + dynamic operation identification. That is, when the type of easily confused bacteria is the same as a certain micro-difference analysis model in the model library, the operation method is the same as (1); when the type of easily confused bacteria is different from the micro-difference analysis model in the model library, the easily confused bacteria are evaluated, that is, whether the easily confused bacteria comply with the dynamic operation start-up rules in the preset dynamic operation start-up rule library is judged; if it does not comply with the dynamic operation start-up rules, the original matching result is directly returned, and prompts and instructions are given for the easily confused bacteria in the original result; if it complies with the dynamic operation start-up rules, the dynamic matching algorithm is used to perform dynamic operation identification on the easily confused bacteria to obtain the dynamic operation identification result.

[0127] The establishment process of the dynamic operation startup rule base is as follows: for the easily confused bacteria generated during the MALDI-TOF MS identification process, the easily confused identification species are activated, and the MALDI-TOF MS method is used for multiple identification evaluations. The identification process needs to follow the following rules: use multi-factor orthogonal experiments such as different treatment methods, different experienced personnel operations, different reagent batches, and different instrument states for evaluation, and include the evaluation results in the dynamic operation startup rule base. The evaluation results have the following situations: 1. The identification results of all test species of the same species in the orthogonal experiment are consistent, the highest scoring species is the correct result, and the difference in identification scores between the correct species and the wrong species in the same report is greater than 3%. When distinguishing between the several easily confused bacteria used for testing, only the dynamic algorithm can be used for final distinction; such easily confused bacteria can be included in the dynamic operation rule base 2. The identification results of all test species of the same species in the orthogonal experiment are inconsistent, but after a limited number of dynamic operations to eliminate the wrongly identified species, the remaining identification results can match the model startup rules. This type of situation can be finally distinguished through dynamic calculation + model method, and this type of easily confused bacteria can be included in the dynamic calculation rule base; 3. The identification results of all test species of the same species in the orthogonal test are inconsistent, and there is an incorrect identification of the highest-scoring species. In this case, easily confused bacteria cannot be dynamically calculated and should be excluded from the dynamic calculation startup rule base; 4. The identification results of all test species of the same species in the orthogonal test are excluded after a dynamic calculation. The easily confused bacteria in this case cannot be dynamically calculated and should be excluded from the dynamic calculation startup rule base; 5. Unevaluated easily confused bacteria should be excluded from the dynamic calculation startup rule base.

[0128] (4) Directly perform multiple rounds of dynamic calculation identification. When there are easily confused bacteria in the species identification of the original matching result (i.e., mass spectrometry identification result), the first round of easily confused bacteria identification is started. If after the first round of easily confused bacteria identification, the easily confused bacteria that were incorrectly identified in the species identification are removed (i.e., after the first round of easily confused bacteria identification, only one species score is greater than the minimum tolerance value of species-level identification), the final species identification result is given and the identification report output is completed. If the first round of identification of easily confused bacteria cannot completely remove the easily confused bacteria that were incorrectly identified in species (i.e., after the first round of identification of easily confused bacteria, there are at least two species with scores greater than the minimum tolerance value for species-level identification), the dynamic operation will continue until the species identification leaves a unique species name (i.e., after multiple dynamic operations, only the unique species identification score remains greater than the minimum tolerance value for species-level identification), and then the final identification result will be reported; or after these multiple rounds of dynamic operations, all easily confused species in the original matching results are reduced to genus identification (i.e., the scores of all species-identified bacteria are reduced to below the minimum tolerance value for species identification after multiple rounds of dynamic operations), indicating that the dynamic operation cannot make a final distinction between the easily confused bacteria. At this time, the original matching results will be returned, and prompts and instructions will be given for the easily confused species in the original results.

[0129] (5) Dynamic judgment + dynamic calculation identification. When there are easily confused bacteria in the species identification of the original matching results, evaluate whether the strains can be distinguished using only dynamic calculations. If the easily confused bacteria species match the rules in the dynamic calculation startup rule library that can complete the final identification using only dynamic calculations, the original matching results are dynamically calculated (the calculation method is the same as in (4), and will not be repeated here) to give the final identification result. When there are easily confused bacteria in the species identification of the original matching results, after evaluation, the dynamic calculation cannot be started, the original matching results are returned, and prompts and instructions are given for the easily confused bacteria species in the original results.

[0130] The dynamic matching algorithm is based on the original database matching results, and it dynamically compares the spectra of a given strain in a small range again in the results, removes the results with large deviations, and improves the accuracy of the results. However, it is worth noting that the accuracy of the results is improved in the dynamic matching process, while losing some accuracy to a certain extent, that is, in the dynamic operation process, the spectrum score will decrease at a certain proportion with the number of operations. Therefore, it is necessary to reasonably control the number of dynamic operations, so as to improve the accuracy of the results while ensuring the correctness of the results. The model library is to conduct micro-difference analysis on each type of strain that is easily confused, determine their methods of micro-difference analysis (including eigenvalue analysis and spectrum trend analysis), and establish the differentiation method in the model library. When the model library contains a micro-difference analysis model with the same report result as the original database matching method, a secondary analysis of the model is performed to further analyze the easily confused bacteria and give a definite result.

[0131] In other embodiments of the present invention, for the highest identification score at the genus level, and mixed identification results of different genera at the genus level, the different genera can also be accurately distinguished based on the above method to obtain a unique genus identification result at the genus level.

[0132] The innovations of this application are as follows: (1) By improving the MALDI-TOF MS algorithm for microbial identification, the system's ability to accurately identify easily confused bacteria is improved; (2) The dynamic matching algorithm and micro-difference analysis model used in this application can be directly applied to the accurate distinction of easily confused bacteria by MALDI-TOF MS, and the accuracy does not vary from person to person, which improves the scalability of the method.

[0133] See also Figure 4 As shown, the embodiment of the present invention discloses a device for identifying easily confusable bacteria, which may specifically include:

[0134] The judgment module 11 is used to obtain the mass spectrometry identification result sent by the original mass spectrometer, and judge whether there are easily mixed bacteria in the mass spectrometry identification result. If there are easily mixed bacteria in the mass spectrometry identification result, it is judged whether there is a micro-difference analysis model of the same type as the easily mixed bacteria in the preset model library;

[0135] The model analysis module 12 is used for, if there is a micro-difference analysis model of the same type as the easily mixed bacteria in the preset model library, calling the micro-difference analysis model to analyze and identify the easily mixed bacteria, so as to obtain a model analysis and identification result;

[0136] The dynamic operation and identification module 13 is used to use a dynamic matching algorithm to perform dynamic operation and identification on the easily confused bacteria if there is no micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, or to use a dynamic matching algorithm and call the micro-difference analysis model to perform dynamic operation and identification on the easily confused bacteria to obtain a dynamic operation and identification result.

[0137] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the method for identifying easily confusable bacteria performed by the electronic device disclosed in any of the aforementioned embodiments.

[0138] In this embodiment, the power supply 23 is used to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0139] In addition, the memory 22, as a carrier for storing resources, can be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon include an operating system 221, a computer program 222 and data 223, etc. The storage method can be temporary storage or permanent storage.

[0140] Among them, the operating system 221 is used to manage and control the hardware devices and computer programs 222 on the electronic device 20 to realize the operation and processing of the data 223 in the memory 22 by the processor 21, which can be Windows, Unix, Linux, etc. In addition to including a computer program that can be used to complete the easily mixed bacteria identification method performed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 can further include a computer program that can be used to complete other specific tasks. In addition to including data transmitted from an external device received by the easily mixed bacteria identification device, the data 223 can also include data collected by its own input and output interface 25, etc.

[0141] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0142] Furthermore, an embodiment of the present application also discloses a computer-readable storage medium, in which a computer program is stored. When the computer program is loaded and executed by a processor, the steps of the method for identifying easily confusable bacteria disclosed in any of the aforementioned embodiments are implemented.

[0143] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants 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, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0144] The above is a detailed introduction to the method, device, equipment and storage medium for identifying easily confused bacteria provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for identifying easily confusable bacteria, characterized in that: include: Obtaining the mass spectrometry identification result sent by the original mass spectrometer, determining whether there are easily confused bacteria in the mass spectrometry identification result, and if there are easily confused bacteria in the mass spectrometry identification result, determining whether there is a micro-difference analysis model of the same type as the easily confused bacteria in the preset model library; If there is a micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, the micro-difference analysis model is retrieved to analyze and identify the easily confused bacteria to obtain a model analysis and identification result; If there is no micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, a dynamic matching algorithm is used to perform dynamic calculation identification on the easily confused bacteria, or a dynamic matching algorithm is used and the micro-difference analysis model is called to perform dynamic calculation identification on the easily confused bacteria to obtain a dynamic calculation identification result.

2. The method for identifying easily confusable bacteria according to claim 1, characterized in that: Before determining whether there is a micro-difference analysis model of the same type as the easily mixed bacteria in the preset model library, the method further includes: The micro-difference analysis method was used to construct each micro-difference analysis model; Each of the micro-difference analysis models is saved in the model library.

3. The method for identifying easily confusable bacteria according to claim 2, characterized in that: The method of using the micro-difference analysis method to construct each micro-difference analysis model includes: Acquire historical data, adopt characteristic value analysis method and use preset characteristic value analysis system to analyze the historical data to obtain fingerprints of various strains, process and perform big data analysis on the fingerprints of various strains to construct micro-difference analysis models; Alternatively, a graph trend analysis method and an artificial intelligence learning method are used to autonomously learn the bacterial species types in the historical data to obtain the graph trend differences between different groups, a reference graph trend differentiation model is constructed based on the graph trend differences, and the reference graph trend differentiation model is verified and corrected to obtain each of the micro-difference analysis models.

4. The method for identifying easily confusable bacteria according to claim 1, characterized in that: The micro-difference analysis model includes a chain analysis model; the construction process of the chain analysis model includes: Obtaining a protein fingerprint of a known strain that is easily confused with a mass spectrometer, and performing noise processing on the protein fingerprint to obtain the processed protein fingerprint; The peak values ​​of different species in the processed protein fingerprint map are multiplexed and compared to obtain the special unique peak values ​​corresponding to different species, and the special unique peak values ​​are classified using a preset binary chain analysis model to construct a chain analysis model, and the chain analysis model is trained to obtain a target chain analysis model.

5. The method for identifying easily confusable bacteria according to claim 4, characterized in that: The peak multiple analysis and comparison of different species in the processed protein fingerprint map to obtain the special unique peaks corresponding to different species include: The processed protein fingerprint map is subjected to multiple analyses using a preset typing software, and multiple peak comparisons are performed on different species in the processed protein fingerprint map to obtain the specific unique peaks corresponding to different species.

6. The method for identifying easily confusable bacteria based on the chain analysis model according to claim 4, characterized in that: The training of the chain analysis model to obtain a target chain analysis model includes: Acquire historical species protein fingerprint peaks, and use the historical species protein fingerprint peaks to train the chain analysis model to obtain training results; The parameters in the chain analysis model are modified and adjusted based on the training results to obtain a target chain analysis model.

7. The method for identifying easily confusable bacteria according to claim 1, characterized in that: The process of using the dynamic matching algorithm to perform dynamic calculation and identification on the easily mixed bacteria is as follows: Setting an initial tolerance value for mass spectrum matching, and matching the mass spectrum of easily mixed bacteria to be matched with the mass spectrum database to obtain an initial matching result, an initial peak matching situation, and a database spectrum corresponding to the initial tolerance value; Using the initial matching result and the initial peak matching situation, judging whether the initial tolerance value satisfies a preset convergence condition, and if the initial tolerance value satisfies the preset convergence condition, taking the initial matching result as the dynamic calculation identification result of the easily confusable bacteria; If the initial tolerance value does not meet the preset convergence condition, the initial matching result and the initial peak matching condition are calculated to obtain an optimized tolerance value; The database spectrum is matched based on the optimized tolerance value to obtain the current matching result and the current peak matching situation, and the process of judging the convergence condition is repeatedly executed until the optimized tolerance value satisfies the convergence condition, and the current matching result is used as the dynamic calculation identification result of the easily confused bacteria.

8. The method for identifying easily confusable bacteria according to claim 7, characterized in that: The initial tolerance value for mass spectrum matching is set, and the mass spectrum to be matched of easily mixed bacteria is matched with the mass spectrum database to obtain an initial matching result, an initial peak matching situation and a database spectrum corresponding to the initial tolerance value, including: Adaptively set the initial tolerance value for mass spectrum matching; The mass spectrum to be matched of the easily mixed bacteria is matched with all spectra in the mass spectrum database to obtain an initial matching result, an initial peak matching situation and a database spectrum corresponding to the initial tolerance value; wherein the initial peak matching situation includes the error values ​​of all matching peaks in the mass spectrum database and the mass spectrum peaks to be matched of the mass spectrum to be matched.

9. The method for identifying easily confusable bacteria according to claim 7, characterized in that: The using the initial matching result and the initial peak matching situation to judge whether the initial tolerance value meets a preset convergence condition includes: A termination equation is set, the initial matching result and the initial peak matching condition are input into the termination equation, and it is determined whether the initial tolerance value meets the preset convergence condition; the termination equation is: ； in, is the initial matching result, This is the initial peak matching situation.

10. The method for identifying easily confusable bacteria according to claim 7, characterized in that: The determining whether the initial tolerance value satisfies a preset convergence condition, if the initial tolerance value satisfies the preset convergence condition, includes: Determine whether all mass spectrum peak error values ​​in the initial matching result are not greater than the initial tolerance value; If all mass spectrum peak error values ​​in the initial matching result are not greater than the initial tolerance value, the initial peak matching situation is uniquely identified. If the uniqueness identification passes, it is determined that the initial tolerance value meets the preset convergence condition.

11. The method for identifying easily confusable bacteria according to any one of claims 7 to 10, characterized in that: Before calculating the initial matching result and the initial peak matching situation, the method further includes: Set the intermediate variable matching results and intermediate variable peak matching conditions; The values ​​of the initial matching result and the initial peak matching situation are assigned to the intermediate variable matching result and the intermediate variable peak matching situation.

12. The method for identifying easily confusable bacteria according to claim 11, characterized in that: The calculating the initial matching result and the initial peak matching condition to obtain an optimized tolerance value includes: A tolerance adjustment equation is set, and the intermediate variable matching result and the intermediate variable peak matching situation are input into the tolerance adjustment equation for calculation to obtain an optimized tolerance value; the tolerance adjustment equation is: ； in, To optimize the tolerance value, is the intermediate variable matching result, This is the peak matching situation of the intermediate variable.

13. The method for identifying easily confusable bacteria according to claim 1, characterized in that: The method of using a dynamic matching algorithm to dynamically calculate and identify the easily mixed bacteria includes: Determine whether there is a target dynamic operation start rule corresponding to each dynamic operation start rule in the dynamic operation start rule base; If there is a target dynamic operation start rule corresponding to each of the dynamic operation start rules in the dynamic operation start rule library, the target dynamic operation start rule is started, and a dynamic matching algorithm is used to perform dynamic operation identification on easily confusable bacteria.

14. The method for identifying easily confusable bacteria according to claim 13, characterized in that: Before determining whether there is a target dynamic operation activation rule corresponding to each dynamic operation activation rule in the dynamic operation activation rule library, the method further includes: Build dynamic operation startup rules based on business needs and in accordance with the preset rule building method; Each of the dynamic operation activation rules is saved in the preset dynamic operation activation rule library.

15. The method for identifying easily confusable bacteria according to claim 1, characterized in that: The method of using the dynamic matching algorithm and calling the micro-difference analysis model to perform dynamic calculation identification on the easily mixed bacteria to obtain the dynamic calculation identification result includes: Setting an initial tolerance value for mass spectrum matching, performing a first round of dynamic calculation, matching the mass spectrum of easily mixed bacteria to be matched with the mass spectrum database, so as to obtain an initial matching result, an initial peak matching situation, and a database spectrum corresponding to the initial tolerance value; Using the initial matching result and the initial peak matching situation, it is judged whether the initial tolerance value satisfies the preset convergence condition. If the initial tolerance value satisfies the preset convergence condition, the dynamic operation ends, and the initial matching result is used as the dynamic operation identification result of the easily confusable bacteria; If the initial tolerance value does not meet the preset convergence condition, that is, the first round of dynamic calculation cannot obtain an identification result, then determine whether there is a micro-difference analysis model of the same type as the current matching result in the preset model library; if there is a micro-difference analysis model of the same type as the current matching result in the preset model library, then call the micro-difference analysis model to analyze and identify the current matching result to obtain a model analysis and identification result; If there is no micro-difference analysis model of the same type as the current matching result in the preset model library, the initial matching result and the initial peak matching situation are calculated to obtain an optimized tolerance value; A new round of dynamic calculation and model matching process is repeated based on the optimized tolerance value until an identification result is obtained.

16. A device for identifying easily confusable bacteria, characterized in that: include: A judgment module is used to obtain the mass spectrometry identification result sent by the original mass spectrometer, and judge whether there are easily mixed bacteria in the mass spectrometry identification result. If there are easily mixed bacteria in the mass spectrometry identification result, it is judged whether there is a micro-difference analysis model of the same type as the easily mixed bacteria in the preset model library; A model analysis module is used for, if there is a micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, calling the micro-difference analysis model to analyze and identify the easily confused bacteria, so as to obtain a model analysis and identification result; The dynamic operation and identification module is used to use a dynamic matching algorithm to perform dynamic operation and identification on the easily confused bacteria if there is no micro-difference analysis model of the same type as the easily confused bacteria in the preset model library, or to use a dynamic matching algorithm and call the micro-difference analysis model to perform dynamic operation and identification on the easily confused bacteria to obtain a dynamic operation and identification result.

17. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor is used to execute the computer program to implement the method for identifying easily confusable bacteria as described in any one of claims 1 to 15.

18. A computer-readable storage medium, characterized in that: Used to store a computer program; wherein, when the computer program is executed by a processor, the method for identifying easily confusable bacteria as described in any one of claims 1 to 15 is implemented.

Images