A method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry

Abstract

The present invention discloses a method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry, and relates to the technical field of aquatic product detection. The method comprises: adding an internal standard solution and acetonitrile according to a predetermined detection scheme to obtain a target mixed solution; performing a drug residue detection on multiple sample solutions based on liquid chromatography-tandem mass spectrometry; collecting growth characteristic data of target samples to be tested; inputting a drug residue prediction model to predict drug residues, and outputting a second detection result; performing deviation calculation to obtain a first detection deviation. The method solves the technical problems that the existing drug residue detection in aquatic products is difficult to detect trace drug residues in aquatic products with high sensitivity and selectivity, and is unable to comprehensively and accurately evaluate the drug residue situation in aquatic products, which leads to low accuracy, sensitivity, and efficiency of drug residue detection, and achieves the technical effect of shortening the detection time and improving the efficiency and accuracy of drug residue detection.

Description

Technical Field

[0001] The present application relates to the technical field related to aquatic product detection, and specifically to a method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry. Background Art

[0002] With the rapid development of aquatic product breeding industry, the problem of drug residues has received increasing attention from the society. Drug residues not only affect the food safety of aquatic products, but may also pose a potential threat to human health. Efficient and accurate drug residue detection methods are of great significance to ensure the safety of aquatic products. Among them, liquid chromatography-tandem mass spectrometry (LC-MS) technology has become the main means of drug residue detection in aquatic products due to its high separation efficiency, high selectivity and high sensitivity. Liquid chromatography-tandem mass spectrometry combines the high separation ability of liquid chromatography (HPLC) with the high sensitivity and selectivity of mass spectrometry (MS) to form a powerful separation and analysis technology. In the detection of drug residues in aquatic products, LC-MS technology can effectively separate a variety of drug residues and perform qualitative and quantitative analysis through characteristic ion mass chromatograms. However, liquid chromatography-tandem mass spectrometry is difficult to detect trace drug residues in aquatic products with high sensitivity and selectivity, and cannot comprehensively and accurately evaluate the drug residues in aquatic products. It also increases the detection time, thereby affecting the efficiency, accuracy and reliability of drug residue detection in aquatic products.

[0003] Therefore, in the current technologies related to drug residue detection in aquatic products, it is difficult to detect trace drug residues in aquatic products with high sensitivity and selectivity, and it is impossible to comprehensively and accurately evaluate the drug residue situation in aquatic products, which leads to technical problems such as low accuracy, sensitivity and efficiency of drug residue detection. Summary of the invention

[0004] The present application provides a method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry, and adopts technical means such as chemical extraction, chromatographic separation and mass spectrometry detection to solve the technical problems existing in the existing detection of drug residues in aquatic products, namely, difficulty in detecting trace drug residues in aquatic products with high sensitivity and high selectivity, inability to comprehensively and accurately evaluate the drug residue situation in aquatic products, and thus resulting in low accuracy, sensitivity and efficiency of drug residue detection, thereby achieving the technical effect of shortening the detection time and improving the efficiency and accuracy of drug residue detection.

[0005] The present application provides a method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry, the method comprising: crushing a target sample to be tested and placing it in a centrifuge tube, adding an internal standard solution and acetonitrile according to a predetermined detection scheme to obtain a target mixed solution, centrifugally oscillating, fractionating, nitrogen blowing and filtering the target mixed solution to obtain multiple sample solutions that meet expected indicators; based on liquid chromatography-tandem mass spectrometry, performing a drug residue detection on the multiple sample solutions once, and determining a first detection result according to a retention time deviation and a relative ion abundance deviation; collecting growth characteristic data of the target sample to be tested, wherein the growth characteristic data includes sample attribute data, water environment data and drug addition data; inputting the sample attribute data, water environment data and drug addition data into a drug residue prediction model to perform drug residue prediction, and outputting a second detection result; performing deviation calculation on the first detection result and the second detection result to obtain a first detection deviation, and if the first detection deviation meets a predetermined deviation threshold, outputting the first detection result as a drug residue detection result.

[0006] In a possible implementation, the output of the drug residue detection result performs the following processing: counting the drug types whose first detection deviation does not meet the predetermined deviation threshold, and obtaining a plurality of suspected drug types; under the same conditions of a single detection, performing a secondary detection on the target test sample based on the plurality of suspected drug types, and obtaining a third detection result; performing deviation calculation on the first detection result and the third detection result, and obtaining a second detection deviation; if the second detection deviation meets the predetermined deviation threshold, performing a mean calculation on the first detection result and the third detection result of the suspected drug type, and outputting the drug residue detection result.

[0007] In a possible implementation, before the primary drug residue test is performed, the following processing is performed: a predetermined drug detection type is obtained, wherein the predetermined drug detection type includes amides, sulfonamides, quinolones, malachite green and diazepam; based on the amides, sulfonamides, quinolones, malachite green and diazepam, a standard stock solution satisfying a predetermined purity is configured according to a predetermined scheme to obtain a plurality of standard solutions, and a primary drug residue test is performed based on the plurality of standard solutions, wherein the predetermined purity is a component content greater than 98%.

[0008] In a possible implementation, the determination of the first test result also performs the following processing: configuring liquid chromatography parameters and mass spectrometry parameters according to the predetermined drug detection type; based on liquid chromatography-tandem mass spectrometry, according to the liquid chromatography parameters and mass spectrometry parameters, taking the multiple standard solutions as a reference, performing a drug residue test on the multiple sample solutions once, calculating the retention time deviation and relative ion abundance deviation between the sample solution and the standard solution, and determining multiple retention time deviations and multiple relative ion abundance deviations; judging the presence of the analyte according to the multiple retention time deviations and the multiple relative ion abundance deviations, and determining multiple drug residue types, wherein if the retention time deviation meets a predetermined time difference threshold and the relative ion abundance deviation meets a predetermined abundance difference threshold, it indicates the presence of drug residues; based on the multiple drug residue types, calculating the residue content according to the drug content calculation formula, determining multiple drug residue contents, and outputting the first test result according to the multiple drug residue types and the multiple drug residue contents.

[0009] In a possible implementation, the residual content is calculated according to the drug content calculation formula, multiple drug residual contents are determined, and the following processing is performed: The expression of the drug content calculation formula is:

[0010]

[0011] Wherein, w is the content of the component to be tested in the aquatic product, C is the concentration of the component to be tested in the standard solution, m is the mass of the test sample, V is the fixed volume, V1 is the sub-volume, and V2 is the extraction volume.

[0012] In a possible implementation, the output of the second test result also performs the following processing: based on the predetermined drug test type, feature extraction is performed on the sample attribute data, water environment data and drug addition data according to drug residue associated indicators to obtain multiple associated data sets; the multiple associated data sets are input into multiple residue prediction units of a drug residue prediction model to perform drug residue prediction, and multiple second drug residue types and multiple second residue contents are output, and the second test result is generated according to the multiple second drug residue types and multiple second residue contents.

[0013] In a possible implementation, the method of obtaining the drug residue association index further performs the following processing: randomly selecting a first drug detection type from the predetermined drug detection types; taking the sample attribute characteristics of the target sample to be tested and the first drug detection type as constraints, performing a correlation analysis on water environment characteristics and drug addition characteristics to determine multiple feature correlations; selecting a feature indicator corresponding to a feature correlation greater than a predetermined correlation threshold, setting it as a first residue association indicator, and constructing the drug residue association index based on the mapping of the first drug detection type and the first residue association indicator.

[0014] In a possible implementation, the construction of the first residue prediction unit also performs the following processing: taking the sample attribute characteristics and the first residue association index as constraints, big data searches historical breeding records to obtain a retrieval data set, wherein the retrieval data includes product attribute characteristics, residue association indicators and drug residue detection results; performing difference dimensionality reduction on the retrieval data set according to a predetermined dimensionality reduction algorithm to obtain a standard data set, using the standard data set to perform supervised training and verification training on the BP neural network to obtain a first residue prediction branch that meets the convergence conditions; and generating the first residue prediction unit based on the first drug detection type and the first residue prediction branch.

[0015] In a possible implementation, the standard data set is obtained and the following processing is also performed: a first retrieval data is randomly selected from the retrieval data set, wherein the first retrieval data includes a first product attribute feature and a first residual association index; based on the first product attribute feature, similarity calculations are performed on other product attribute features in the retrieval data set, and the number of times the similarity is less than a predetermined product similarity threshold is counted, which is recorded as a first random coefficient; if the first random coefficient is greater than the predetermined threshold, based on the first residual association index, similarity calculations are performed on other residual association indicators in the retrieval data set, and the number of times the similarity is less than a predetermined indicator similarity threshold is counted, which is recorded as a second random coefficient; if the second random coefficient is greater than the predetermined threshold, the first retrieval data is added to the standard data set.

[0016] The invention proposes a method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry, wherein an internal standard solution and acetonitrile are added according to a predetermined detection scheme to obtain a target mixed solution; based on liquid chromatography-tandem mass spectrometry, a drug residue detection is performed on multiple sample solutions once; growth characteristic data of the target sample to be tested is collected; a drug residue prediction model is input to predict drug residues, and a second detection result is output; deviation calculation is performed to obtain a first detection deviation, thereby solving the technical problems existing in the existing drug residue detection in aquatic products, such as difficulty in detecting trace drug residues in aquatic products with high sensitivity and high selectivity, inability to comprehensively and accurately evaluate the drug residue situation in aquatic products, and thus resulting in low accuracy, sensitivity and efficiency of drug residue detection, thereby achieving the technical effect of shortening the detection time and improving the efficiency and accuracy of drug residue detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solution of the embodiment of the present disclosure, the accompanying drawings of the embodiment of the present disclosure will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0018] Figure 1 A schematic flow chart of a method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry provided in an embodiment of the present application;

[0019] Figure 2 A schematic diagram of a process for determining a first detection result in a method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry provided in an embodiment of the present application. DETAILED DESCRIPTION

[0020] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0021] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0022] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0023] The present application provides a method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry, such as Figure 1 As shown, the method includes:

[0024] The liquid chromatography-tandem mass spectrometry detection method is used to determine the residues of amide alcohols, sulfonamides, quinolones, malachite green and diazepam in aquatic products. The principle is to extract the amide alcohols, sulfonamides, quinolones, malachite green and diazepam in the test sample with acetonitrile, take out nitrogen blown and then redissolve (malachite green needs to be purified by solid phase extraction), filter membrane, and after determination, use the internal standard method (diazepam uses the external standard method) for quantification. It combines the separation ability of liquid chromatography with the selectivity and sensitivity of mass spectrometry, and can detect multiple drug residues at the same time. Among them, for malachite green, a solid phase extraction purification step is usually required to remove interfering substances and improve the accuracy and sensitivity of the analysis. The internal standard method is a quantitative analysis method, in which a known amount of internal standard substance is added to each sample to be tested, and the test substance is quantified by comparing the response intensity ratio of the test substance and the internal standard substance. The internal standard substance is extracted, purified and analyzed together with the sample to be tested. The concentration of the sample to be tested is calculated by calculating the peak area or peak height ratio of the sample to be tested and the internal standard, combined with the known concentration of the internal standard. The external standard method is a quantitative analysis method, in which a standard curve is prepared using a standard substance of known concentration, and then the target substance in the sample to be tested is quantified by comparing the response intensity of the sample to be tested with the standard curve. During the analysis process, the standard substance and the sample to be tested are injected separately, the standard substance is used to establish the standard curve, and the sample to be tested is used for quantitative analysis.

[0025] Step S100, crush the target sample to be tested and place it in a centrifuge tube, add the internal standard solution and acetonitrile according to the predetermined detection scheme, obtain the target mixed solution, centrifuge, extract, blow nitrogen and filter the target mixed solution, and obtain multiple sample solutions that meet the expected indicators. Use appropriate tools (such as blenders, homogenizers) to crush the target sample to be tested (such as fish, shrimp and other aquatic products), and then place it in a clean centrifuge tube, according to the predetermined detection scheme, add a certain amount of internal standard solution and an appropriate amount of acetonitrile as an extractant to the sample in the centrifuge tube, centrifuge, extract, blow nitrogen and filter, wherein the internal standard solution is usually used to correct errors in the experimental process and improve the accuracy of the test results. Acetonitrile is a commonly used organic solvent that can effectively extract drug residues in the sample. Specifically, the centrifuge tube to which acetonitrile and the internal standard solution are added is sealed and placed on an oscillator for oscillation, so that acetonitrile is fully mixed with the sample to improve the extraction efficiency of drug residues. After the oscillation is completed, the centrifuge tube is placed in a centrifuge for centrifugation to separate the solid particles in the solution from the liquid. After centrifugation, carefully take out the supernatant, which contains the drug residue extracted by acetonitrile and the internal standard solution. Repeat the extraction of the residue to improve the recovery rate of the drug residue. Transfer the supernatant obtained to an appropriate container and use a nitrogen blower to blow nitrogen to separate the acetonitrile layer, that is, remove the acetonitrile by nitrogen blowing (i.e., evaporating under a nitrogen flow) and concentrate the residue. During the nitrogen blowing process, the flow rate and temperature of the nitrogen need to be controlled to avoid damage or loss of the drug residue. The concentrated solution after nitrogen blowing may contain some solid particles or impurities, which need to be removed by filtration, for example, using an appropriate filter membrane (such as a regenerated cellulose filter membrane) for filtration. After the filtration is completed, multiple sample solutions that meet the expected indicators are finally obtained for subsequent liquid chromatography-tandem mass spectrometry detection.

[0026] Step S200, based on liquid chromatography-tandem mass spectrometry, the multiple sample solutions are tested for drug residues once, and the first test result is determined according to the retention time deviation and the relative ion abundance deviation. Using liquid chromatography-tandem mass spectrometry, multiple sample solutions are tested for drug residues once. Specifically, the obtained multiple sample solutions are injected into a liquid chromatograph, separated by a chromatographic column, and the separated target test samples enter a mass spectrometer, where they are ionized into ions at the ion source, and characteristic fragment ions are generated through a multi-stage fragmentation process of a tandem mass spectrometer. Qualitative and quantitative analysis is performed based on the retention time and relative ion abundance of these characteristic fragment ions, and then the first test result is determined based on the retention time deviation and the relative ion abundance deviation, wherein the retention time deviation refers to the difference between the retention time of the target test sample in the sample solution and the relative ion abundance deviation of the target sample in the standard working solution (or a reference solution of known concentration) under the same chromatographic conditions. The difference between the retention times of the compounds should be considered. If the retention time of the analyte in the sample deviates from that of the standard substance within ±2.5%, it indicates that the analyte in the sample and the standard substance are the same substance; relative ion abundance refers to the proportion or intensity of a specific ion in all detected ions in mass spectrometry analysis. It is used to describe the relative relationship between the signal intensity of a specific ion and the sum of the signal intensities of all ions in the sample under given conditions. The relative abundance of the qualifier ions of each component in the sample is compared with the relative abundance of the corresponding qualifier ions in the standard solution with similar concentrations. If the relative abundance of the qualifier ions in the sample deviates from that in the standard solution within the range specified in the following table:

[0027] Relative ion abundance ＞50％ 20％～50％ 10％～20％ ≤10％ Maximum allowed deviation ±20％ ±25％ ±30％ ±50％

[0028] It can be considered that the corresponding analyte exists in the sample, and this is ultimately used as the first test result of drug residue detection.

[0029] In a possible implementation, step S200 further includes step S201, obtaining a predetermined drug detection type, wherein the predetermined drug detection type includes amide alcohols, sulfonamides, quinolones, malachite green and diazepam. The contents of chloramphenicol, florfenicol and thiamphenicol in the amide alcohols were all greater than 98.0%; the contents of sulfonamides were sulfadiazine, sulfadoxine, sulfamerazine, sulfadimethoxine, sulfamethoxazole, sulfaisoxazole, sulfamethixadiazole, sulfachloropyridazine, sulfathiazole, sulfamethoxine and sulfaquinoxaline were all greater than 98.0%; the contents of ciprofloxacin hydrochloride, enrofloxacin, norfloxacin, ofloxacin, lomefloxacin and pefloxacin in the quinolones were all greater than 98.0%; the contents of malachite green oxalate and leucomalachite green in malachite green were all greater than 98.0%; the contents of diazepam in diazepam were all greater than 98.0%.

[0030] Step S202, based on the amide alcohols, sulfonamides, quinolones, malachite green and diazepam, standard stock solutions meeting the predetermined purity are configured according to a predetermined scheme to obtain multiple standard solutions, and a drug residue test is performed based on the multiple standard solutions, wherein the predetermined purity is a component content greater than 98%. ① Take an appropriate amount of amide alcohols, sulfonamides and diazepam standards (equivalent to about 10 mg of each active ingredient), accurately weigh them, dissolve them in methanol and make up to the volume in a 10 mL volumetric flask, and prepare a standard stock solution of 1 mg / mL; ② Take an appropriate amount of quinolone standards (equivalent to about 10 mg of each active ingredient), accurately weigh them, add 0.2 mL of formic acid, dissolve them in methanol and make up to the volume in a 10 mL volumetric flask, and prepare a standard stock solution of 1 mg / mL; ③ Take an appropriate amount of malachite green oxalate and leucomalachite green standards (equivalent to about 1 mg of each active ingredient), dissolve them in acetonitrile and make up to the volume in a 10 mL volumetric flask, and prepare a standard stock solution of 100 mg / L. Dilute the standard stock solutions ①, ②, and ③ step by step to 1.00 mg / L standard intermediate solutions; prepare the amide alcohol mixed standard intermediate solution into a 0.0500 ng / mL to 2.00 ng / mL standard series solution with 20% acetonitrile aqueous solution; prepare the quinolone and sulfonamide mixed standard intermediate solution into a 1.00 ng / mL to 20.0 ng / mL standard series solution with 20% acetonitrile aqueous solution; prepare the malachite green mixed standard intermediate solution into a 0.200 ng / mL to 10.0 ng / mL standard series solution with acetonitrile + 0.1% formic acid + 5mmoL / L ammonium acetate solution; prepare diazepam with 20% acetonitrile aqueous solution. The standard intermediate solution was prepared into a standard series of solutions ranging from 0.500 ng / mL to 10.0 ng / mL. An appropriate amount of deuterated chloramphenicol was accurately weighed and prepared into a 100 mg / L stock solution with methanol. An appropriate amount of deuterated sulfa-2-methoxazole and deuterated sulfa-2-methoxazole were accurately weighed and prepared into a 100 mg / L stock solution with methanol. Deuterated norfloxacin, deuterated ciprofloxacin and deuterated enrofloxacin were added with 0.2 mL of formic acid and prepared into a 100 mg / L stock solution with methanol. An appropriate amount of deuterated malachite green and deuterated crypto-malachite green were accurately weighed and prepared into a 100 mg / L stock solution with acetonitrile. The internal standard stock solution was diluted step by step into a 1.00 mg / L standard intermediate solution.

[0031] In one possible implementation, Figure 2 As shown, step S200 further includes step S210, configuring liquid chromatography parameters and mass spectrometry parameters according to the predetermined drug detection type. Liquid chromatography parameters include chromatographic column, mobile phase, flow rate, column temperature, injection volume, for example, for amide alcohols, chromatographic column: C18 column (2.6μm, 4.6mm×100mm), flow rate 800μL / min, column temperature 40°C, injection volume 20μL, mobile phase, A: water, B: acetonitrile, gradient elution program is as follows:

[0032] Time (min) Mobile phase A (water) / % Mobile phase B (acetonitrile) / % ----- 75 25 3.0 35 65 3.1 10 90 4.5 10 90 4.6 75 25 6.2 75 25

[0033] The mass spectrometry parameters of amide alcohol drugs are as follows:

[0034]

[0035] Step S220, based on liquid chromatography-tandem mass spectrometry, according to the liquid chromatography parameters and mass spectrometry parameters, with the multiple standard solutions as the benchmark, the multiple sample solutions are tested for drug residues once, the retention time deviation and relative ion abundance deviation between the sample solutions and the standard solutions are calculated, and multiple retention time deviations and multiple relative ion abundance deviations are determined. Using liquid chromatography-tandem mass spectrometry, based on the determined multiple standard solutions, according to the liquid chromatography parameters and mass spectrometry parameters, multiple sample solutions are tested for drug residues once, the retention time deviation and relative ion abundance deviation between the sample solutions and the standard solutions are calculated, specifically, the retention time deviation refers to the difference between the retention time of the substance to be tested in the sample and the retention time of the corresponding substance in the standard working solution, the relative ion abundance deviation refers to the difference between the relative ion abundance of the qualitative ion of the substance to be tested in the sample and the relative ion abundance of the corresponding ion in the standard working solution, the retention time deviation calculation formula is:

[0036] Δt i =t sample,i -t standard,i ;

[0037] Among them, Δt i is the retention time deviation of the i-th target sample to be tested, t sample,i is the retention time of the i-th target sample in the sample solution, t standard,i is the retention time of the i-th target sample in the standard solution;

[0038] Relative ion abundance calculation formula:

[0039]

[0040] Among them, ΔI i is the relative ion abundance deviation of the i-th target sample to be tested, I sample,i is the ion abundance of the i-th target compound in the sample solution, I standard,i is the ion abundance of the i-th target sample to be tested in the standard solution; and finally the retention time deviations and multiple relative ion abundance deviations corresponding to multiple drug residue types are determined.

[0041] Step S230, judging the existence of the analyte according to the plurality of retention time deviations and the plurality of relative ion abundance deviations, and determining a plurality of drug residue types, wherein if the retention time deviation satisfies a predetermined time difference threshold and the relative ion abundance deviation satisfies a predetermined abundance difference threshold, it indicates the existence of drug residues. Judging the existence of the analyte according to the plurality of retention time deviations and the plurality of relative ion abundance deviations, and determining a plurality of drug residue types, if the retention time deviation of a sample solution satisfies a predetermined time difference threshold (such as ±0.5 minutes) and the relative ion abundance deviation also satisfies a predetermined abundance difference threshold (such as ±20%), it is judged that the sample contains corresponding drug residues.

[0042] Step S240, based on the multiple drug residue types, calculate the residue content according to the drug content calculation formula, determine multiple drug residue contents, and output the first detection result according to the multiple drug residue types and the multiple drug residue contents.

[0043] Furthermore, step S240 also includes that the drug content calculation formula is expressed as:

[0044]

[0045] Wherein, w is the content of the component to be tested in the aquatic product, C is the concentration of the component to be tested in the standard solution, m is the mass of the test sample, V is the fixed volume, V1 is the fraction volume, and V2 is the extraction volume. The residual content of multiple drug residue types is calculated using the drug content calculation formula to obtain the drug residue content corresponding to each drug residue type, determine multiple corresponding drug residue contents, and output multiple drug residue types and corresponding multiple drug residue contents as the first test result.

[0046] Step S300, collecting growth characteristic data of the target sample to be tested, wherein the growth characteristic data includes sample attribute data, water environment data and drug addition data. Collecting growth characteristic data of the target sample to be tested specifically refers to determining various characteristics and parameters related to the target sample to be tested during its growth, development or cultivation process, which is crucial for understanding the growth environment, growth conditions, physiological state and possible influences (such as drug residues, pests and diseases, etc.) of the sample. Growth characteristic data includes sample attribute data, water environment data and drug addition data. Specifically, sample attribute data refers to various characteristic parameters related to the target sample to be tested itself, including but not limited to variety information (such as fish, shrimp, crab and other species and their varieties), size and weight (such as body length, weight and other sample sizes), gender and age (age is estimated through growth rings, scale size and other indicators), etc.; water environment data refers to water environment parameters that affect the growth of aquatic products, including but not limited to water temperature (affects the growth and metabolism of aquatic products, and different types of aquatic products have different adaptability ranges to water temperature), solubility Oxygen (affects the respiration of aquatic products, and thus affects their growth rate and quality), pH value (affects the growth and survival of aquatic products, and the appropriate pH value range is conducive to the healthy growth of aquatic products), salinity (affects the osmotic pressure and physiological function of aquatic products), water quality indicators (such as ammonia nitrogen, nitrite, nitrate and other water quality indicators, reflecting the degree of water pollution and the health of the breeding environment), etc.; drug addition data refers to information such as the type, dosage, time and method of drugs used in the process of aquatic product breeding, including but not limited to drug types (such as antibiotics, vaccines, growth promoters, etc.), dosage (to assess whether it meets safety standards and usage specifications), use time (drug addition time and duration, so as to analyze the relationship between drug residues and breeding cycle), use method (such as oral, injection, immersion, etc. and frequency, so as to understand the distribution and metabolism of drugs in aquatic products), etc.

[0047] Step S400, input the sample attribute data, water environment data and drug addition data into the drug residue prediction model to perform drug residue prediction, and output the second test result. The sample attribute data, water environment data and drug addition data are input into the drug residue prediction model to perform drug residue prediction, wherein the drug residue prediction model includes multiple residue prediction units, each prediction unit may predict different types of drug residues, specifically, using sample attribute characteristics and residue association indicators as constraints, and using big data retrieval technology to find similar cases or data sets in historical breeding records, in order to obtain more training data to improve the accuracy and generalization ability of the model, and then perform differential dimensionality reduction processing on the retrieved data set according to a predetermined dimensionality reduction algorithm (such as principal component analysis, linear discriminant analysis, etc.), in order to reduce the dimension of the data set while retaining the most critical information to obtain a standard data set, and use the standard data set to obtain the standard data set. The BP neural network is supervised and verified based on the data set. During the training process, the model will continuously adjust its parameters to minimize the prediction error until the convergence condition is met (that is, the prediction error no longer decreases significantly). Based on the trained BP neural network and the specific drug detection type, multiple corresponding drug residue prediction branches are generated. These branches will be used as part of the drug residue prediction model, specifically for predicting specific types of drug residues. Finally, multiple drug residue prediction branches are integrated to form a drug residue prediction model. When sample attribute data, water environment data and drug addition data are input into the drug residue prediction model, the drug residue prediction result is output as the second test result, including multiple drug residue types and corresponding residue content prediction results.

[0048] In a possible implementation, step S400 further includes step S410, based on the predetermined drug detection type, the sample attribute data, water environment data and drug addition data are feature extracted according to the drug residue association index to obtain multiple associated data sets. Drug residue association index refers to various characteristics or factors closely related to drug residue detection, and these indicators may come from multiple aspects such as sample attribute data, water environment data and drug addition data. Specifically, according to the drug residue association index, feature extraction is performed on the collected sample attribute data, water environment data and drug addition data, that is, information or features useful for drug residue detection are extracted from the original data, and multiple data sets closely related to drug residue detection are obtained, which respectively represent the association between sample attributes, water environment and drug addition and drug residue.

[0049] Step S420, input the multiple associated data sets into the multiple residue prediction units of the drug residue prediction model to perform drug residue prediction, output multiple second drug residue types and multiple second residue contents, and generate the second test result according to the multiple second drug residue types and multiple second residue contents. The multiple associated data sets obtained by feature extraction are used as input and respectively sent to the multiple corresponding residue prediction units of the drug residue prediction model, each prediction unit outputs the corresponding second drug residue type and second residue content according to the input feature data, forming a set of multiple second drug residue types and multiple second residue contents, and the second drug residue types and residue contents output by each prediction unit are aggregated to form a complete prediction result set, thereby generating a second test result, which will include all predicted drug residue types and their corresponding residue contents.

[0050] In a possible implementation, step S410 further includes step S411, randomly selecting a first drug detection type from the predetermined drug detection types. Step S412, taking the sample attribute characteristics of the target sample to be tested and the first drug detection type as constraints, performing correlation analysis on the water environment characteristics and the drug addition characteristics to determine multiple feature correlations. A first drug detection type is randomly selected from the predetermined drug detection types, with the sample attribute characteristics of the target sample to be tested and the selected first drug detection type as constraints. The sample attribute characteristics include at least aquatic product type, growth stage and gender, which are used to guide the correlation analysis between water environment characteristics and drug addition characteristics. Under the constraints, a correlation analysis is performed on the water environment characteristics and drug addition characteristics to evaluate how the water environment characteristics and drug addition characteristics are associated with the first drug detection type and the sample attribute characteristics of the target sample to be tested. Through the correlation analysis, we can determine which water environment characteristics (such as water quality parameters, water temperature, pH value, etc.) and which drug addition characteristics (such as drug type, dosage, usage time, number of additions, etc.) are significantly correlated with the first drug detection type and sample attribute characteristics, and finally determine multiple feature correlations.

[0051] Step S413, select a feature index corresponding to a feature correlation greater than a predetermined correlation threshold, set it as a first residue correlation index, and construct the drug residue correlation index based on the mapping between the first drug detection type and the first residue correlation index. Screen out feature correlations greater than the predetermined correlation threshold from all feature correlations, and the feature index corresponding to the feature correlation greater than this threshold will be considered to be highly correlated with the first drug detection type. Set the feature index corresponding to the feature correlation greater than the predetermined correlation threshold as the first residue correlation index, which represents the water environment characteristics and drug addition characteristics most relevant to the first drug detection type. Construct a drug residue correlation index based on the mapping relationship between the first drug detection type and the first residue correlation index. Specifically, comprehensively consider the characteristics of the first drug detection type and the water environment characteristics and drug addition characteristics highly correlated thereto, and use it for subsequent drug residue prediction, risk assessment or regulatory decision-making, etc.

[0052] In a possible implementation, step S420 further includes step S421, using the sample attribute characteristics and the first residue association index as constraints, searching historical breeding records with big data to obtain a retrieval data set, wherein the retrieval data includes product attribute characteristics, residue association indexes and drug residue detection results; step S422, performing difference dimensionality reduction on the retrieval data set according to a predetermined dimensionality reduction algorithm to obtain a standard data set, and using the standard data set to perform supervised training and verification training on the BP neural network to obtain a first residue prediction branch that meets the convergence conditions; step S423, generating the first residue prediction unit based on the first drug detection type and the first residue prediction branch. Taking the sample attribute characteristics and the first residue association index as constraints, relevant data sets are retrieved from the big data of historical breeding records. The retrieved data sets include product attribute characteristics (such as breeding conditions, feed types, breeding time, etc.), residue association indexes (indicators that may be related to drug residues, such as the content of certain biomarkers) and drug residue test results (actually detected drug residues); a predetermined dimensionality reduction algorithm is used to perform differential dimensionality reduction on the retrieved data sets to reduce the complexity of the data, and finally a more compact and easy-to-process standard data set is obtained; the standard data set is used to perform supervised training and verification training on the BP (back propagation) neural network. Through training, the neural network can learn the relationship between the sample product attribute characteristics and the residue association index and the drug residue test results. When the neural network meets the convergence condition, the training is completed and a model branch that can predict the drug residue is obtained, namely the first residue prediction branch; the trained first residue prediction branch is combined with a specific drug detection type to form a prediction unit specifically used to predict the drug residue of this type. This residue prediction unit contains prediction logic and parameters for a specific drug detection type, can receive the relevant attribute characteristics and residue association index of the breeding product as input, and output the predicted residue of a specific drug in this type of product.

[0053] In a possible implementation, step S421 further includes step S4211, randomly selecting the first retrieval data from the retrieval data set, wherein the first retrieval data includes the first product attribute feature and the first residual association index. Step S4212, taking the first product attribute feature as a benchmark, respectively calculating the similarity of the other product attribute features in the retrieval data set, and counting the number of times the similarity is less than a predetermined product similarity threshold, which is recorded as a first random coefficient. Randomly select a piece of data from the retrieval data set as the first retrieval data, the first retrieval data includes the first product attribute feature and the first residual association index, and taking the first product attribute feature as a benchmark, using Euclidean distance, cosine similarity, etc., respectively calculate the similarity of the attribute features of other products in the retrieval data set, and then counting the number of times the similarity is less than the predetermined product similarity threshold, which reflects how many products have attribute features that are significantly different from the attribute features of the first retrieval data. For example, if Euclidean distance is used, the similarity calculation formula is:

[0054]

[0055] Among them, A i and B i is the value of the i-th dimension of the two product attribute features. For each calculated similarity value, determine whether it is less than the predetermined product similarity threshold. If the similarity value is less than the threshold, increase the count, and count the number of times the similarity is less than the predetermined threshold as the first random coefficient. The formula is expressed as:

[0056] The first random coefficient = ∑(Sim(A, B)<product similarity threshold);

[0057] Among them, Sim(A,B) represents the similarity calculation result between the first product attribute feature and the other product attribute features.

[0058] Step S4213: If the first random coefficient is greater than a predetermined threshold, similarity calculations are performed on other residual association indicators in the retrieval data set based on the first residual association indicator, and the number of times the similarity is less than the predetermined indicator similarity threshold is counted and recorded as the second random coefficient. Step S4214: If the second random coefficient is greater than a predetermined threshold, the first retrieval data is added to the standard data set. If the first random coefficient is greater than the predetermined threshold, it means that the product attribute characteristics of the first retrieval data are relatively unique in the retrieval data set, that is, whether the difference with other data is significant. Taking the first residual association index as the benchmark, the residual association indexes of other data in the retrieval data set are respectively calculated for similarity, and whether the residual association index of the first retrieval data set is also unique enough is further evaluated. The number of times the similarity is less than the predetermined index similarity threshold is counted, which reflects how many residual association indicators are significantly different from the residual association indicators of the first retrieval data. The number of times is recorded as the second random coefficient; if the second random coefficient is greater than the predetermined threshold, it means that the residual association index of the first retrieval data set is also relatively unique in the retrieval data set. When the above two conditions are met (the first random coefficient and the second random coefficient are both greater than the predetermined threshold), the first retrieval data with both unique product attribute characteristics and unique residual association indicators are added to the standard data set to ensure that the data in the standard data set is both representative and unique, which helps to improve the accuracy and effectiveness of residual prediction model training.

[0059] Step S500, performing deviation calculation on the first detection result and the second detection result to obtain a first detection deviation, and if the first detection deviation meets a predetermined deviation threshold, outputting the first detection result as a drug residue detection result. Each drug residue type and residue content in the first test result is compared with the corresponding data in the second test result, and the difference between them is calculated to obtain a first test deviation. If the first test deviation meets the predetermined deviation threshold, the first test result is output as a drug residue test result, wherein the first test deviation includes deviation values ​​corresponding to multiple drug test types in aquatic products, and the predetermined deviation threshold is a standard value pre-set based on historical data, which is used to determine whether the difference between the first test result and the second test result is within an acceptable range. The absolute difference between two independent determination results obtained under repeatability conditions does not exceed 15% of the arithmetic mean. Specifically, if the first test deviation meets the predetermined deviation threshold (the first test deviation is less than or equal to the predetermined deviation threshold), the first test result is recognized and used as the drug residue test result; if the first test deviation does not meet the predetermined deviation threshold (the first test deviation is greater than the predetermined deviation threshold), the one with a large deviation is calibrated, and the experimental test is performed again under exactly the same conditions. If the difference between the first test result and the re-test result is less than the predetermined threshold, the average value of the two test results is selected as the final drug residue test result, thereby reducing the number of repeated experimental tests, saving the drug residue test time, and thus improving the efficiency of drug residue detection in aquatic products.

[0060] In a possible implementation, step S500 further includes step S510, counting the drug types whose first detection deviation does not meet the predetermined deviation threshold, and obtaining a plurality of suspected drug types. Counting all drug types whose first detection deviation does not meet the predetermined deviation threshold, forming a list of suspected drug types, identifying the drug types whose results are uncertain or may be inaccurate in the first detection, and obtaining a plurality of suspected drug types. Step S520, under the same conditions of the first detection, based on the plurality of suspected drug types, performing a secondary detection on the target sample to be tested, and obtaining a third detection result. Under the same conditions of the first detection (such as temperature, humidity, instrument settings, etc.), but only configuring the detection method for the suspected drug type, such as adjusting the detection parameters, using a more sensitive detection reagent, or adopting a different detection strategy, a more accurate secondary detection is performed on the target sample to be tested based on the plurality of suspected drug types, and the detection result is obtained as the third detection result. Step S530, the first test result and the third test result are subjected to deviation calculation to obtain a second test deviation. If the second test deviation satisfies a predetermined deviation threshold, the first test result and the third test result of the suspected drug type are subjected to mean calculation, and the drug residue test result is output. The residual content of the suspected drug type in the first test result is compared with the corresponding data in the third test result, and the difference between them is calculated to obtain a second test deviation. The second test deviation is compared with a predetermined deviation threshold. If the second test deviation satisfies the predetermined deviation threshold (i.e., the deviation is within an acceptable range), the first test result and the third test result are considered to be reliable for the suspected drug type. For the suspected drug type, the first test result and the third test result are subjected to mean calculation to obtain a more accurate residual content estimate, and then the estimate is output as the final drug residue test result.

[0061] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry, characterized in that: include: The target sample to be tested is crushed and placed in a centrifuge tube, an internal standard solution and acetonitrile are added according to a predetermined detection scheme to obtain a target mixed solution, and the target mixed solution is centrifugally shaken, fractionated, nitrogen-blown and filtered to obtain multiple sample solutions that meet the expected indicators; Based on liquid chromatography-tandem mass spectrometry, the multiple sample solutions are tested for drug residues once, and a first test result is determined according to a retention time deviation and a relative ion abundance deviation; Collecting growth characteristic data of the target sample to be tested, wherein the growth characteristic data includes sample attribute data, water environment data and drug addition data; Inputting the sample attribute data, water environment data and drug addition data into a drug residue prediction model to perform drug residue prediction, and outputting a second test result; Calculating the deviation between the first detection result and the second detection result to obtain a first detection deviation, and if the first detection deviation meets a predetermined deviation threshold, outputting the first detection result as a drug residue detection result; Among them, the drug residue test is carried out once, and before that, it also includes: Obtaining a predetermined drug test type, wherein the predetermined drug test type includes amides, sulfonamides, quinolones, malachite green, and diazepam; Based on the amide alcohols, sulfonamides, quinolones, malachite green and diazepam, a standard stock solution meeting a predetermined purity is configured according to a predetermined scheme to obtain multiple standard solutions, and a drug residue test is performed based on the multiple standard solutions, wherein the predetermined purity is a component content greater than 98%; Wherein, outputting the second detection result includes: Based on the predetermined drug detection type, feature extraction is performed on the sample attribute data, water environment data and drug addition data according to drug residue correlation indicators to obtain multiple correlation data sets; The multiple associated data sets are input into multiple residue prediction units of a drug residue prediction model to perform drug residue prediction, multiple second drug residue types and multiple second residue contents are output, and the second detection result is generated according to the multiple second drug residue types and multiple second residue contents.

2. The method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry according to claim 1, characterized in that: Output drug residue test results, including: counting the types of drugs whose first detection deviation does not meet a predetermined deviation threshold, to obtain a plurality of suspected drug types; Under the same conditions of the first test, based on the multiple suspected drug types, a second test is performed on the target test sample to obtain a third test result; The deviation between the first detection result and the third detection result is calculated to obtain a second detection deviation. If the second detection deviation meets a predetermined deviation threshold, the first detection result and the third detection result of the suspected drug type are averaged and the drug residue detection result is output.

3. The method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry according to claim 1, characterized in that: Determining a first test result includes: According to the predetermined drug detection type, configuring liquid chromatography parameters and mass spectrometry parameters; Based on liquid chromatography-tandem mass spectrometry, according to the liquid chromatography parameters and mass spectrometry parameters, taking the multiple standard solutions as a reference, performing a drug residue detection on the multiple sample solutions once, calculating the retention time deviation and relative ion abundance deviation between the sample solutions and the standard solutions, and determining multiple retention time deviations and multiple relative ion abundance deviations; The presence of the analyte is judged according to the multiple retention time deviations and the multiple relative ion abundance deviations, and multiple drug residue types are determined, wherein if the retention time deviation meets a predetermined time difference threshold and the relative ion abundance deviation meets a predetermined abundance difference threshold, it indicates the presence of drug residue; Based on the multiple drug residue types, the residue content is calculated according to the drug content calculation formula to determine multiple drug residue contents, and the first detection result is output according to the multiple drug residue types and the multiple drug residue contents.

4. The method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry according to claim 3, characterized in that: The drug content calculation formula is expressed as: Wherein, w is the content of the component to be tested in the aquatic product, C is the concentration of the component to be tested in the standard solution, m is the mass of the test sample, V is the fixed volume, V1 is the sub-volume, and V2 is the extraction volume.

5. The method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry according to claim 1, characterized in that: Obtain drug residue related indicators, including: randomly selecting a first drug test type from among the predetermined drug test types; Taking the sample attribute characteristics of the target sample to be tested and the first drug detection type as constraints, a correlation analysis is performed on the water environment characteristics and the drug addition characteristics to determine the correlations of multiple characteristics; A feature index corresponding to a feature correlation degree greater than a predetermined correlation threshold is selected as a first residue correlation index, and the drug residue correlation index is constructed based on the first drug detection type and the first residue correlation index mapping.

6. The method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry according to claim 5, characterized in that: Constructing a first residual prediction unit, including: Taking the sample attribute characteristics and the first residue correlation index as constraints, the big data searches historical breeding records to obtain a search data set, wherein the search data includes product attribute characteristics, residue correlation indexes and drug residue test results; Performing difference dimensionality reduction on the retrieval data set according to a predetermined dimensionality reduction algorithm to obtain a standard data set, and using the standard data set to perform supervised training and verification training on the BP neural network to obtain a first residual prediction branch that meets the convergence condition; The first residual prediction unit is generated based on the first drug detection type and a first residual prediction branch.

7. The method for detecting drug residues in aquatic products by liquid chromatography-tandem mass spectrometry according to claim 6, characterized in that: Get the standard data set, including: Randomly selecting first search data from the search data set, wherein the first search data includes a first product attribute feature and a first residual association index; Taking the first product attribute feature as a benchmark, similarity calculations are performed on other product attribute features in the search data set, and the number of times the similarity is less than a predetermined product similarity threshold is counted, which is recorded as a first random coefficient; If the first random coefficient is greater than a predetermined threshold, similarity calculations are performed on other residual association indicators in the retrieval data set based on the first residual association indicator, and the number of times the similarity is less than the predetermined indicator similarity threshold is counted and recorded as the second random coefficient; If the second random coefficient is greater than a predetermined threshold, the first search data is added to the standard data set.

Images