Amino acid biomarker for distinguishing wild silver carp and cultured silver carp, method and application
The amino acid biomarkers were screened out through an amino acid automatic analyzer and a logistic regression model was established, which solved the problem of distinguishing wild and raising silver carp, and improved the scientificity and effectiveness of resource management and protection.
Patent Information
- Application Number
- CN202510244223.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-03
- Publication Date
- 2025-08-05
- Estimated Expiration
- 2045-03-03
AI Technical Summary
The existing technology lacks effective methods to distinguish wild and cultivated silver carp, which leads to overfishing of wild resources and makes it difficult to scientifically manage and protect.
The muscle samples of wild and farmed silver were measured by automatic amino acid analyzer, and amino acid biomarkers were screened out through amino acid spectrum analysis, such as aspartic acid, alanine, phenylalanine, histidine, and arginine, and the logistic regression analysis model was used to determine it, and a discriminant model was established to accurately distinguish wild and farmed silver.
It provides scientific basis, improves supervision of illegal fishing behavior, reduces overfishing of wild resources, and achieves better resource management and protection.
Smart Images
Figure CN120427918A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of silver carp identification or confirmation, and specifically to amino acid biomarkers, methods and applications for distinguishing wild and farmed silver carp. Background Art
[0002] Silver carp is an important aquaculture species in my country, primarily found in the Yangtze River basin. Developing technologies to identify or confirm the presence of wild and farmed silver carp, identifying and promoting biomarkers for both species, and promoting their application can help reduce overfishing of wild stocks. Summary of the Invention
[0003] This application addresses the problem of similar biological traits between wild and farmed silver carp and the lack of identification technology. For the first time, an amino acid automatic analyzer was used to measure and collect muscle samples from different wild and farmed silver carp populations. Amino acid profile analysis was performed to explore the common differences and changes in amino acids between wild and farmed silver carp, and amino acid biomarkers were screened out: aspartic acid (CAS: 56-84-8), alanine (CAS: 56-41-7), phenylalanine (CAS: 63-91-2), histidine (CAS: 71-00-1), and arginine (CAS: 74-79-3). Using the results of the amino acid automatic analyzer test, statistical and logistic regression analysis were performed to obtain a logistic regression analysis model. The logistic regression model was used to discriminate the source of the test samples, and it was able to accurately discriminate between wild and farmed silver carp with good sensitivity and specificity. Furthermore, the amino acid biomarkers and uses provided in this application can provide scientific basis and technical support for relevant law enforcement agencies, strengthen the supervision and crackdown on illegal fishing in the Yangtze River Basin, reduce overfishing of wild resources, and provide a technical basis for better management and protection of silver carp resources.
[0004] To this end, the embodiments of the present application disclose at least the following technical solutions:
[0005] In a first aspect, the embodiments disclose a biomarker comprising at least one of aspartic acid, alanine, phenylalanine, histidine, and arginine.
[0006] In the second aspect, the embodiment discloses a method for distinguishing wild silver carp from farmed silver carp. The method includes: obtaining a first logistic regression model and a first threshold; obtaining multiple relative content percentages of multiple biomarker contents in the test sample, where the multiple biomarkers are aspartic acid, alanine, phenylalanine, histidine, and arginine; inputting the multiple relative content percentages into the first logistic regression model to obtain a first wild probability, where the first wild probability represents the probability that the test sample is wild silver carp; and determining whether the test sample is the wild silver carp or the farmed silver carp based on the size of the first wild probability and the first threshold. Wherein, the first logistic regression model is: X = -2.707×A+2.776×B-1.281×C-2.114×D+5.323×E-29.477; first wild probability = 1 / (1+e －X ); wherein A, B, C, D, and E are the relative content percentages of aspartic acid, alanine, phenylalanine, histidine, and arginine, respectively.
[0007] In some embodiments of the second aspect, if the first wild probability is greater than the first threshold, the test sample is wild silver carp. If the first wild probability is less than the first threshold, the test sample is farmed silver carp.
[0008] In an embodiment of the second aspect, the step of determining the first threshold value includes: using the relative content percentages of aspartic acid, alanine, phenylalanine, histidine, and arginine in the collected muscle samples as a training set; training a binary logistic regression model with the training set to obtain multiple first wild probability values; calculating the sensitivity (true positive rate) and 1-specificity (false positive rate) of each first wild probability value, and plotting an ROC curve with 1-specificity as the horizontal axis and sensitivity as the vertical axis; calculating the Youden index corresponding to each first wild probability (Youden index = sensitivity + specificity - 1), and obtaining the maximum Youden index, the first wild probability corresponding to the maximum Youden index being the first threshold value. In some embodiments, the first threshold value is 0.1928.
[0009] In the third aspect, the embodiment discloses a method for distinguishing wild silver carp from farmed silver carp. The method includes obtaining a second logistic regression model and a second threshold; obtaining multiple relative content percentages of multiple biomarker contents in the test sample, where the multiple biomarkers are aspartic acid, alanine, histidine, and arginine; inputting the multiple relative content percentages into the second logistic regression model to obtain a second wild probability, where the second wild probability represents the probability that the test sample is wild silver carp; determining whether the test sample is the wild silver carp or the farmed silver carp based on the size of the second wild probability and the second threshold. Wherein, the second logistic regression model is: X = -2.769×A+2.938×B-2.378×C+5.239×D-34.888; the second wild probability = 1 / (1+e －X ); wherein A, B, C, and D are the relative content percentages of aspartic acid, alanine, histidine, and arginine, respectively.
[0010] In some embodiments of the third aspect, if the second wild probability is greater than the second threshold, the test sample is wild silver carp. If the second wild probability is less than the second threshold, the test sample is farmed silver carp.
[0011] In an embodiment of the third aspect, the step of determining the second threshold value includes: using the relative content percentages of aspartic acid, alanine, histidine, and arginine in the collected muscle samples as a training set; training a binary logistic regression model with the training set to obtain multiple second wild probability values; calculating the sensitivity (true positive rate) and 1-specificity (false positive rate) of each second wild probability value, and plotting an ROC curve with 1-specificity as the horizontal axis and sensitivity as the vertical axis; calculating the Youden index corresponding to each second wild probability (Youden index = sensitivity + specificity - 1), and obtaining the maximum Youden index, the second wild probability corresponding to the maximum Youden index is the second threshold value. In some embodiments, the second threshold value is 0.1795.
[0012] In the fourth aspect, the embodiment discloses a method for distinguishing wild silver carp from farmed silver carp. The method includes: obtaining a third logistic regression model and a third threshold; obtaining multiple relative content percentages of multiple biomarker contents in the test sample, where the multiple biomarkers are alanine, histidine, and arginine; inputting the multiple relative content percentages into the third logistic regression model to obtain a third wild probability, where the third wild probability characterizes the probability that the test sample is wild silver carp; determining whether the test sample is the wild silver carp or the farmed silver carp based on the size of the third wild probability and the third threshold. Wherein, the third logistic regression model is: X = 2.494×A-2.157×B+7.298×C-74.351; third wild probability = 1 / (1+e －X); A, B, and C are the relative content percentages of alanine, histidine, and arginine, respectively.
[0013] In some embodiments of the fourth aspect, if the third wild probability is greater than the third threshold, the test sample is wild silver carp. If the third wild probability is less than the third threshold, the test sample is farmed silver carp.
[0014] In some embodiments of the fourth aspect, the step of determining the third threshold value includes: using the relative content percentages of alanine, histidine, and arginine in the collected muscle samples as a training set; training a binary logistic regression model with the training set to obtain multiple third wild probability values; calculating the sensitivity (true positive rate) and 1-specificity (false positive rate) of each third wild probability value, and plotting an ROC curve with 1-specificity as the horizontal axis and sensitivity as the vertical axis; calculating the Youden index corresponding to each third wild probability (Youden index = sensitivity + specificity - 1), and obtaining the maximum Youden index, the third wild probability corresponding to the maximum Youden index being the third threshold value. In some embodiments, the third threshold value is 0.2855.
[0015] In the fifth aspect, the embodiment discloses a method for distinguishing wild silver carp from farmed silver carp. The method includes: obtaining a fourth logistic regression model and a fourth threshold; obtaining multiple relative content percentages of multiple biomarker contents in the test sample, where the multiple biomarkers are alanine and arginine; inputting the multiple relative content percentages into the fourth logistic regression model to obtain a fourth wild probability, where the fourth wild probability represents the probability that the test sample is wild silver carp; determining whether the test sample is the wild silver carp or the farmed silver carp based on the size of the fourth wild probability and the fourth threshold. Wherein, the fourth logistic regression model is: X=4.334×A+7.664×B-97.780; fourth wild probability=1 / (1+e －X ); A and B are the relative content percentages of alanine and arginine, respectively.
[0016] In some embodiments of the fifth aspect, if the fourth wild probability is greater than the fourth threshold, the test sample is wild silver carp. If the fourth wild probability is less than the fourth threshold, the test sample is farmed silver carp.
[0017] In some embodiments of the fifth aspect, the step of determining the fourth threshold value includes: using the relative percentages of alanine and arginine in the collected muscle samples as a training set; training a binary logistic regression model with the training set to obtain a plurality of fourth wild probability values; calculating the sensitivity (true positive rate) and 1-specificity (false positive rate) of each fourth wild probability value, and plotting an ROC curve with 1-specificity as the abscissa and sensitivity as the ordinate; calculating the Youden index corresponding to each fourth wild probability (Youden index = sensitivity + specificity - 1), and obtaining the maximum Youden index, the fourth wild probability corresponding to the maximum Youden index being the fourth threshold value. In some embodiments, the fourth threshold value is 0.3020.
[0018] In a sixth aspect, the embodiments disclose the use of the biomarkers described in the first aspect and / or the reagents for detecting the biomarkers described in the first aspect in distinguishing wild and farmed silver carp. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] Figure 1 The confusion matrix test results of the amino acid biomarker composition (aspartic acid, alanine, phenylalanine, histidine, arginine) provided in the example were used to distinguish wild silver carp from farmed silver carp.
[0020] Figure 2 The confusion matrix test results of the amino acid biomarker composition (aspartic acid, alanine, histidine, arginine) provided in the example were used to distinguish wild silver carp from farmed silver carp.
[0021] Figure 3 The confusion matrix test results of the amino acid biomarker composition (alanine, histidine, arginine) provided in the example were used to distinguish wild silver carp from farmed silver carp.
[0022] Figure 4 The confusion matrix test results of the amino acid biomarker composition (alanine, arginine) provided in the example were used to distinguish wild silver carp from farmed silver carp. DETAILED DESCRIPTION
[0023] In order to make the purpose, technical solutions and advantages of this application more clearly understood, the present application is further described in detail below with reference to the following examples. It should be understood that the specific examples described herein are merely for the purpose of explaining this application and are not intended to limit this application. Reagents not described in detail in this application are all conventional reagents and can be obtained from commercial channels; methods not specifically described in detail are all conventional experimental methods and can be obtained from the prior art.
[0024] Mining of amino acid biomarkers for distinguishing wild and farmed silver carp
[0025] 1. Collection of muscle samples from silver carp
[0026] The wild silver carp populations involved in the examples were sourced from the Minjiang and Jialingjiang Rivers of the Yangtze River, the Three Gorges Reservoir area, Dongting Lake, and Poyang Lake. The farmed silver carp populations involved in the examples were sourced from farms in Hubei, Hunan, and Jiangsu provinces. Samples were obtained from the upper back muscles of silver carp and immediately frozen in liquid nitrogen after collection. They were then stored in a -80°C freezer until amino acid extraction was performed.
[0027] 2. Determination and analysis of amino acid content in samples
[0028] Amino acid determination follows the standard GB 5009.124-2016, "National Food Safety Standard - Determination of Amino Acids in Food." This method can measure the content of 16 amino acids. Approximately 0.02 g of dried muscle sample was weighed and placed in a hydrolysis tube. 10 mL of 6 mol / L hydrochloric acid was added, the tube was vacuum-sealed, and hydrolysis was performed at 110°C for 22 h. After cooling, the volume was adjusted to 25 mL with pure water. 5 mL was then diluted to 10 mL. Finally, 5 mL was transferred to a small weighing bottle and incubated in an 80°C waterbath until dry. 1 mL of deionized water was added and the sample was incubated in an 80°C waterbath until dry. Repeat this three times. Dissolve the remaining material in the weighing bottle with 5 mL of 0.02 mol / L hydrochloric acid, transfer to a centrifuge tube, and mix thoroughly. 1 mL was filtered through a 0.22 μm aqueous filter into a sample vial. Equal volumes of the mixed amino acid standard working solution and sample solution were injected into the amino acid analyzer. The amino acid concentrations in the sample solution were calculated using the external standard method based on peak area.
[0029] The absolute content of each amino acid in the sample solution was calculated according to the formula Calculation, where c i A is the content of amino acid i in the sample assay solution, in nanomoles per milliliter (nmol / mL); i A is the peak area of amino acid i in the sample determination solution; s is the peak area of amino acid s in the amino acid standard working solution; c s is the content of amino acid s in the amino acid standard working solution, in nanomoles per milliliter (nmol / mL). The absolute content of each amino acid in the sample is calculated according to the formula Calculation, where Xi is the content of amino acid i in the sample, in grams per 100 grams (g / 100g); ci is the content of amino acid i in the sample solution, in nanomoles per milliliter (nmol / mL); F is the dilution factor; V is the volume of the sample hydrolyzate transferred to the fixed volume, in milliliters (mL); M is the relative molecular mass of amino acid i, in grams per mole (g / mol). The names and relative molecular masses of various amino acids are shown in Table 1; m is the sample weight, in grams (g); 10 9 The coefficient for converting the sample content from nanograms (ng) to grams (g); 100 is the conversion coefficient. The relative percentage of each amino acid in the sample is calculated according to the formula Where Y iis the relative content percentage of amino acid i in the sample; X i is the absolute content of amino acid i in the sample.
[0030] 3. Statistical analysis methods
[0031] A t-test was used to calculate significant differences in amino acid data between different wild and farmed silver carp populations. The relative percentage content of each amino acid was entered into SPSS for the construction and analysis of a binary logistic regression model. The relative percentage content of each amino acid was presented as the mean plus or minus the standard deviation (SD). An asterisk (*) indicates a statistically significant difference between the experimental and control groups. Specifically, *P < 0.05 indicates a significant difference, and **P < 0.01 indicates an extremely significant difference.
[0032] 4. Results
[0033] Comparisons of silver carp samples from wild and farmed populations were performed, with a P value of < 0.05 used to screen for significantly different amino acids. Analysis revealed five significantly different amino acids: aspartic acid, alanine, phenylalanine, histidine, and arginine. The relative percentages and P values for these amino acids are shown in Table 1.
[0034] Table 1 Comparison of amino acid biomarkers between wild and farmed silver carp
[0035]
[0036] Amino acid biomarker composition for distinguishing wild silver carp from farmed silver carp and its use
[0037] The mainstay of current data discrimination models is to construct models using multiple indicators for comprehensive judgment. Logistic regression, as a simple and intuitive model, is easy to understand and interpret, requiring no complex mathematical derivations or calculations while maintaining high computational efficiency. Furthermore, logistic regression provides highly interpretable output, enabling classification by setting thresholds. The model's coefficients and intercept provide information on the degree and direction of a variable's contribution to the target variable.
[0038] 1. Selection of amino acid biomarker compositions
[0039] This example tests the amino acid content in muscle samples of wild and farmed silver carp, and trains a binary logistic regression model based on the test results. Amino acids with high statistical significance, low measurement error, and high data integrity are selected as biomarkers for distinguishing wild from farmed silver carp. The amino acid biomarkers provided in this example for distinguishing wild from farmed silver carp include aspartic acid, alanine, phenylalanine, histidine, and arginine.
[0040] 2. How to distinguish wild silver carp from farmed silver carp
[0041] The embodiments also disclose a method for distinguishing wild silver carp from farmed silver carp. The method includes: obtaining a first logistic regression model and a first threshold; obtaining multiple relative content percentages of multiple biomarkers in a test sample; inputting the multiple relative content percentages into the first logistic regression model to obtain a first wild probability; and determining whether the test sample is wild or farmed silver carp based on the difference between the first wild probability and the first threshold. The multiple biomarkers include aspartic acid, alanine, phenylalanine, histidine, and arginine.
[0042] In some embodiments, the step of obtaining a first logistic regression model includes:
[0043] 1) Obtain training samples
[0044] In this embodiment, muscle samples from 30 wild silver carp and 30 farmed silver carp were tested for amino acid content using the same method described above. Peak areas for aspartic acid, alanine, phenylalanine, histidine, and arginine were obtained for each sample. The relative percentages of the target amino acids were calculated using the respective calculation formulas. The relative percentages of aspartic acid, alanine, phenylalanine, histidine, and arginine in these muscle samples served as training samples.
[0045] 2) Training
[0046] A binary logistic regression model is trained using the relative percentages of aspartic acid, alanine, phenylalanine, histidine, and arginine as independent variables and the probability of a sample being wild type as the dependent variable. In some embodiments, the binary logistic regression model training process includes: performing a composition analysis on the dependent and independent variables to determine whether they meet the prerequisites for logistic regression; performing a significance test on the independent variables, including degrees of freedom, significance, and score; performing a Hosmer-Lemmeshow test to verify the applicability and rationality of the model and ensure its accuracy; and training the binary logistic regression model to obtain a first logistic regression model based on the degree of influence of the independent variables on the dependent variable.
[0047] In some steps, the examples performed single-degree-of-freedom tests on the independent variables of aspartic acid, alanine, phenylalanine, histidine, and arginine in muscle samples from wild and farmed silver carp. The results, shown in Table 2, indicate that the levels of these four biomarkers, as independent variables, had high overall statistical scores, with significance levels less than 0.05, indicating that they significantly impacted the corresponding variables overall.
[0048] Table 2 Significance tests and scores of independent variables in the logistic regression model of biomarkers of wild and farmed silver carp
[0049]
[0050] In some steps, the examples performed a Hosmer-Lemmeshow test on logistic regression models of biomarkers in wild and farmed silver carp. Under multiple degrees of freedom, the binary logistic regression model achieved a significance of P = 0.552 > 0.05, thus accepting the null hypothesis. The results, shown in Table 3, demonstrate a good fit with the real data and reliably reflect the relationships between the original variables.
[0051] Table 3 Hosmer-Lemmeshow test for the logistic regression model of biomarkers in wild and farmed silver carp
[0052] step Chi-square degrees of freedom Significance 1 6.862 8 0.552
[0053] 3) Obtain the first logistic regression model
[0054] In some steps, the embodiment performs binary logistic regression analysis on the logistic regression model of biomarkers of wild silver carp and farmed silver carp. Table 4 shows the effect size Exp(B) of the regression model and the significance of each biomarker.
[0055] Table 4 Analysis of binary logistic regression models in biomarker models for wild and farmed silver carp
[0056]
[0057] Example The first logistic regression model was obtained based on the relative content percentages of the above 30 wild silver carp muscle samples and 30 farmed silver carp muscle samples as follows:
[0058] X=-2.707×A+2.776×B-1.281×C-2.114×D+5.323×E-29.477;
[0059] First wild probability = 1 / (1+e －X );
[0060] Among them, A, B, C, D, and E are the relative content percentages of aspartic acid, alanine, phenylalanine, histidine, and arginine, respectively, and the first wild probability represents the probability that the test sample is wild silver carp.
[0061] The steps provided in the embodiment are to obtain a first threshold value by calculating the ROC curve. The steps provided in some embodiments include: using the relative content percentages of aspartic acid, alanine, phenylalanine, histidine, and arginine in muscle samples of 30 wild silver carp and 30 farmed silver carp as a training set; training a binary logistic regression model with the training set to obtain multiple first wild probability values as described above; calculating the sensitivity (true positive rate) and 1-specificity (false positive rate) of each first wild probability value, and plotting the ROC curve with 1-specificity as the abscissa and sensitivity as the ordinate; calculating the Youden index corresponding to each predicted wild probability (Youden index = sensitivity + specificity - 1), and obtaining the maximum Youden index, wherein the first predicted wild probability corresponding to the maximum Youden index is the first threshold value.
[0062] In some embodiments, if the first wild probability is greater than a first threshold, the test sample is determined to be wild silver carp. If the first wild probability is less than the first threshold, the test sample is determined to be farmed silver carp.
[0063] In some embodiments, the ROC curve calculation using the above test set yields a first threshold of 0.1928. That is, when the first wild probability is greater than 0.1928, the test sample is determined to be wild silver carp; when the first wild probability is less than 0.1928, the test sample is determined to be farmed silver carp.
[0064] 3. Identify application and testing
[0065] In some test cases, aspartic acid, alanine, phenylalanine, histidine, and arginine from 58 muscle samples (including 28 wild and 30 farmed) were selected as biomarkers, and their relative content percentages were input into the above-mentioned first logistic regression model to obtain the first wild probability; based on the size of the first wild probability and the first threshold (0.1928), it was determined whether the test sample was wild silver carp or farmed silver carp.
[0066] Binary logistic regression models were constructed using the relative percentages of aspartic acid, alanine, phenylalanine, histidine, and arginine in the training set, respectively. A first logistic regression model was constructed using the combined relative percentages of aspartic acid, alanine, phenylalanine, histidine, and arginine. These 58 muscle samples were tested, and the discriminant results were analyzed using receiver operating characteristic (ROC) curves. The results are shown in Table 5. As shown in Table 5, the first logistic regression model constructed using the combined relative percentages of aspartic acid, alanine, phenylalanine, histidine, and arginine exhibited higher AUC values, sensitivity, and specificity.
[0067] Table 5 ROC test results of biomarkers and their combinations for discriminating wild and farmed silver carp
[0068] Biomarkers AUC Sensitivity Specificity Aspartic acid 0.849 0.714 0.900 Alanine 0.850 1.000 0.667 Phenylalanine 0.857 0.857 0.800 Histidine 0.826 0.821 0.700 Arginine 0.920 0.893 0.800 Composition 0.964 1.000 0.833
[0069] The discrimination results are input into SPSS20.0 analysis software for confusion matrix test. Confusion matrix, also known as error matrix, is a standard format for expressing accuracy evaluation, which is expressed in the form of a matrix with n rows and n columns. Specific evaluation indicators include overall accuracy, mapping accuracy, user accuracy, etc. These accuracy indicators reflect the accuracy of image classification from different aspects. Each column of the confusion matrix represents the predicted category, and the total number of each column represents the number of data predicted to be in that category; each row represents the true category of the data, and the total number of data in each row represents the number of data instances in that category. The results are as follows Figure 1 As shown in the figure, only 5 of the 58 silver carp samples were misjudged, and the overall accuracy of the interactive validation was as high as 91.38%.
[0070] This suggests that aspartic acid, alanine, phenylalanine, histidine, and arginine can be used as biomarkers to differentiate wild and farmed silver carp from the Yangtze River basin. The excellent receiver operating characteristic (ROC) curve, high correct diagnosis rate, and good biomarker complementarity suggest that these five amino acids have the potential to distinguish wild from farmed silver carp.
[0071] Amino acid biomarker composition for distinguishing wild silver carp from farmed silver carp and its use
[0072] 1. Selection of amino acid biomarker compositions
[0073] This example tests the amino acid content in muscle samples of wild and farmed silver carp, and trains a binary logistic regression model based on the test results. Amino acids with high statistical significance, low measurement error, and high data integrity are selected as biomarkers for distinguishing wild from farmed silver carp. The amino acid biomarkers provided in this example include aspartic acid, alanine, histidine, and arginine.
[0074] The embodiments also provide a kit for detecting these biomarkers, which includes reagents for detecting these biomarkers, such as reagents required for automatic amino acid analyzer testing.
[0075] 2. How to distinguish wild silver carp from farmed silver carp
[0076] The embodiments also disclose a method for distinguishing wild silver carp from farmed silver carp. The method includes: obtaining a second logistic regression model and a second threshold; obtaining multiple relative content percentages of multiple biomarkers in a test sample; inputting the multiple relative content percentages into the second logistic regression model to obtain a second wild probability; and determining whether the test sample is wild or farmed silver carp based on the difference between the second wild probability and the second threshold. The multiple biomarkers include aspartic acid, alanine, histidine, and arginine.
[0077] In some embodiments, the step of obtaining a second logistic regression model includes:
[0078] 1) Obtain training samples
[0079] Example: 30 wild silver carp muscle samples and 30 farmed silver carp muscle samples were tested using the same automatic amino acid analyzer described above to obtain the peak areas for aspartic acid, alanine, histidine, and arginine in each sample. The relative content percentages of the target amino acids were calculated using the respective calculation formulas. The relative content percentages of aspartic acid, alanine, histidine, and arginine in these muscle samples served as training samples.
[0080] 2) Training
[0081] A binary logistic regression model is trained using the relative percentages of aspartic acid, alanine, histidine, and arginine as independent variables and the probability of a sample being wild type as the dependent variable. In some embodiments, the training process of the binary logistic regression model includes: performing a composition analysis on the dependent and independent variables to determine whether they meet the prerequisites for logistic regression; performing a significance test on the independent variables, including degrees of freedom, significance, and score; performing a Hosmer-Lemmeshow test to verify the applicability and rationality of the model and ensure its accuracy; and training the binary logistic regression model to obtain a second logistic regression model based on the degree of influence of the independent variables on the dependent variable.
[0082] In some steps, the examples performed single-degree-of-freedom tests on the muscle sample levels of the independent variables aspartate, alanine, histidine, and arginine, including providing degrees of freedom, significance, and score testing. The results, shown in Table 6, show that the overall statistical scores of the levels of these four biomarkers as independent variables were high, with significance levels less than 0.001, indicating that the independent variables had a highly significant overall impact on the corresponding variables.
[0083] Table 6 Significance tests and scores of independent variables in the logistic regression model of biomarkers of wild and farmed silver carp
[0084]
[0085] In some steps, the examples performed a Hosmer-Lemmeshow test on the logistic regression models for biomarkers in wild and farmed silver carp. Under multiple degrees of freedom, the significance of the binary logistic regression model was P = 0.611 > 0.05, meaning the null hypothesis was accepted. The results, shown in Table 7, demonstrate a good fit between the real data and reliably reflect the relationships between the original variables.
[0086] Table 7 Hosmer-Lemmeshow test for the logistic regression model of wild and farmed silver carp
[0087] step Chi-square degrees of freedom Significance 1 6.326 8 0.611
[0088] 3) Obtain the second logistic regression model
[0089] In some steps, the embodiment performs binary logistic regression analysis on the biomarkers for distinguishing wild silver carp from farmed silver carp. Table 8 shows the effect size Exp(B) of the regression model and the significance of each biomarker.
[0090] Table 8 Binary logistic regression analysis of wild and farmed silver carp
[0091]
[0092] Example The first logistic regression model was obtained based on the relative content percentages of the above 30 wild silver carp muscle samples and 30 farmed silver carp muscle samples as follows:
[0093] X=-2.769×A+2.938×B-2.378×C+5.239×D-34.888;
[0094] Second wild probability = 1 / (1+e －X );
[0095] Among them, A, B, C, and D are the relative content percentages of aspartic acid, alanine, histidine, and arginine, respectively, and the second wild probability represents the probability that the test sample is wild silver carp.
[0096] The steps provided in the embodiment are to obtain a second threshold value by ROC curve calculation. The steps provided in some embodiments include: using the relative content percentages of aspartic acid, alanine, histidine, and arginine in muscle samples of 30 wild silver carp and 30 farmed silver carp as a training set; training a binary logistic regression model with the training set to obtain multiple second wild probability values as described above; calculating the sensitivity (true positive rate) and 1-specificity (false positive rate) of each second wild probability value, and plotting the ROC curve with 1-specificity as the abscissa and sensitivity as the ordinate; calculating the Youden index corresponding to each predicted wild probability (Youden index = sensitivity + specificity - 1), and obtaining the maximum Youden index, wherein the first predicted wild probability corresponding to the maximum Youden index is the second threshold value.
[0097] In some embodiments, if the second wild probability is greater than a second threshold, the test sample is determined to be wild silver carp. If the second wild probability is less than the second threshold, the test sample is determined to be farmed silver carp.
[0098] In some embodiments, the ROC curve calculation using the above test set yields a second threshold of 0.1795. That is, when the second wild probability is greater than 0.1795, the test sample is determined to be wild silver carp; when the second wild probability is less than 0.1795, the test sample is determined to be farmed silver carp.
[0099] 3. Identify application and testing
[0100] In some test cases, aspartic acid, alanine, histidine, and arginine from 58 muscle samples (including 28 wild and 30 farmed) were selected as biomarkers, and their relative content percentages were input into the above-mentioned second logistic regression model to obtain the second wild probability; based on the size of the second wild probability and the second threshold (0.1795), it was determined whether the test sample was wild silver carp or farmed silver carp.
[0101] Binary logistic regression models were constructed using the relative percentages of aspartic acid, alanine, histidine, and arginine in the training set, respectively. A second logistic regression model was constructed using the combined relative percentages of aspartic acid, alanine, histidine, and arginine. These 58 muscle samples were tested, and the discrimination results were analyzed using receiver operating characteristic (ROC) curves. The results are shown in Table 9. As shown in Table 9, the second logistic regression model constructed using the combined relative percentages of aspartic acid, alanine, histidine, and arginine for discriminating the test samples exhibited higher AUC values, sensitivity, and specificity.
[0102] Table 9 ROC test results of each biomarker and its combination for discriminating wild and farmed silver carp
[0103] Biomarkers AUC Sensitivity Specificity Aspartic acid 0.849 0.714 0.900 Alanine 0.850 1.000 0.667 Histidine 0.826 0.821 0.700 Arginine 0.920 0.893 0.800 Composition 0.967 1.000 0.833
[0104] The discrimination results were input into SPSS20.0 analysis software for confusion matrix test. The results are as follows: Figure 2 As shown in the figure, only 5 of the 58 silver carp samples were misjudged, and the overall accuracy of the interactive validation was as high as 91.38%.
[0105] This suggests that aspartic acid, alanine, histidine, and arginine can be used as biomarkers to differentiate wild and farmed silver carp from the Yangtze River basin. The excellent receiver operating characteristic (ROC) curve, high correct diagnosis rate, and good biomarker complementarity all indicate that these four amino acids have the potential to distinguish wild from farmed silver carp.
[0106] Amino acid biomarker composition for distinguishing wild silver carp from farmed silver carp and its use
[0107] 1. Selection of amino acid biomarker compositions
[0108] This example tests the amino acid content in muscle samples of wild and farmed silver carp, and trains a binary logistic regression model based on the test results. Amino acids with high statistical significance, low measurement error, and high data integrity are selected as biomarkers for distinguishing wild from farmed silver carp. The amino acid biomarkers provided in this example include alanine, histidine, and arginine.
[0109] The embodiments also provide a kit for detecting these biomarkers, which includes reagents for detecting these biomarkers, such as reagents required for automatic amino acid analyzer testing.
[0110] 2. How to distinguish wild silver carp from farmed silver carp
[0111] The embodiments also disclose a method for distinguishing wild silver carp from farmed silver carp. The method includes: obtaining a third logistic regression model and a third threshold; obtaining multiple relative percentages of multiple biomarker levels in a test sample; inputting the multiple relative percentages into the third logistic regression model to obtain a third probability of wildness; and determining whether the test sample is wild or farmed silver carp based on the difference between the third probability of wildness and the third threshold. The multiple biomarkers include alanine, histidine, and arginine.
[0112] In some embodiments, the step of obtaining the third logistic regression model includes:
[0113] 1) Obtain training samples
[0114] Example: 30 wild silver carp muscle samples and 30 farmed silver carp muscle samples were analyzed using the same automatic amino acid analyzer described above to obtain the peak areas for alanine, histidine, and arginine in each sample. The relative percentages of the target amino acids were calculated using the respective calculation formulas. The relative percentages of alanine, histidine, and arginine in these muscle samples served as training samples.
[0115] 2) Training
[0116] A binary logistic regression model is trained using the relative percentages of alanine, histidine, and arginine as independent variables and the probability of a sample being wild as the dependent variable. In some embodiments, the training process of the binary logistic regression model includes: performing a composition analysis on the dependent and independent variables to determine whether they meet the prerequisites for logistic regression; performing a significance test on the independent variables, including degrees of freedom, significance, and score; performing a Hosmer-Lemmeshow test to verify the applicability and rationality of the model and ensure its accuracy; and training the binary logistic regression model to obtain a third logistic regression model based on the degree of influence of the independent variables on the dependent variable.
[0117] In some steps, the embodiment performed a single degree of freedom test on muscle samples for the independent variables alanine, histidine, and arginine, including providing degrees of freedom, significance, and score testing. As shown in Table 10, the overall statistical score of the independent variables is high, with a significance of less than 0.001, indicating that the independent variables have a very significant impact on the corresponding variables overall.
[0118] Table 10 Significance test and score of independent variables in the logistic regression model of biomarkers of wild and farmed silver carp
[0119]
[0120] To further analyze the impact of the constant and independent variables on the dependent variable in terms of multiple degrees of freedom, the example also performed a Hosmer-Lemmeshow test on the logistic regression model. Under the multiple degrees of freedom condition, the significance of the binary logistic regression model was P = 0.338 > 0.05, i.e., the null hypothesis was accepted. The results, as shown in Table 11, show that the constructed binary logistic regression model fits the real data well and can reliably reflect the reliable relationship between the original variables.
[0121] Table 11 Hosmer-Lemmeshow test for the logistic regression model of wild and farmed silver carp
[0122] step Chi-square degrees of freedom Significance 1 9.057 8 0.338
[0123] 3) Obtain the third logistic regression model
[0124] In some steps, the embodiment performs binary logistic regression analysis on the logistic regression model of biomarkers of wild silver carp and farmed silver carp. Table 12 shows the effect size Exp(B) of the regression model and the significance of each biomarker.
[0125] Table 12 Binary logistic regression analysis of wild and farmed silver carp
[0126]
[0127]
[0128] Example According to the relative content percentages of the muscle samples of 30 wild silver carp and 30 farmed silver carp, a third logistic regression model was obtained as follows:
[0129] X=2.494×A-2.157×B+7.298×C-74.351;
[0130] The probability of the third wild = 1 / (1+e －X );
[0131] A, B, and C are the relative content percentages of alanine, histidine, and arginine, respectively, and the third wild probability represents the probability that the test sample is wild silver carp.
[0132] The steps provided in the embodiment are to obtain a third threshold value by ROC curve calculation. The steps provided in some embodiments include: using the relative content percentages of alanine, histidine, and arginine in muscle samples of 30 wild silver carp and 30 farmed silver carp as a training set; training a binary logistic regression model with the training set to obtain multiple third wild probability values as described above; calculating the sensitivity (true positive rate) and 1-specificity (false positive rate) of each third wild probability value, and plotting the ROC curve with 1-specificity as the abscissa and sensitivity as the ordinate; calculating the Youden index corresponding to each predicted wild probability (Youden index = sensitivity + specificity - 1), and obtaining the maximum Youden index, wherein the third predicted wild probability corresponding to the maximum Youden index is the third threshold value.
[0133] In some embodiments, if the third wild probability is greater than a third threshold, the test sample is determined to be wild silver carp. If the third wild probability is less than the third threshold, the test sample is determined to be farmed silver carp.
[0134] In some embodiments, the third threshold value obtained by calculating the ROC curve using the above test set is 0.2855. That is, when the third wild probability is greater than 0.2855, the test sample is determined to be wild silver carp; when the third wild probability is less than 0.2855, the test sample is determined to be farmed silver carp.
[0135] 3. Identify application and testing
[0136] In some test cases, alanine, histidine, and arginine from 58 muscle samples (including 28 wild and 30 farmed) were selected as biomarkers, and their relative content percentages were input into the above-mentioned third logistic regression model to obtain the third wild probability; based on the size of the third wild probability and the third threshold (0.2855), it was determined whether the test sample was wild silver carp or farmed silver carp.
[0137] Binary logistic regression models were constructed using the relative percentages of alanine, histidine, and arginine in the training set, respectively. A third logistic regression model was constructed using the combined percentages of alanine, histidine, and arginine. These 58 muscle samples were tested, and the discriminant results were analyzed using receiver operating characteristic (ROC) curves. The results are shown in Table 14. As shown in Table 14, the third logistic regression model constructed using the combined percentages of alanine, histidine, and arginine exhibited higher AUC values, sensitivity, and specificity.
[0138] Table 14 ROC test results of each biomarker and its combination for discriminating wild and farmed silver carp
[0139] Biomarkers AUC Sensitivity Specificity Alanine 0.850 1.000 0.667 Histidine 0.826 0.821 0.700 Arginine 0.920 0.893 0.800 Composition 0.954 1.000 0.800
[0140] The discrimination results were input into SPSS20.0 analysis software for confusion matrix test. The results are as follows: Figure 3 As shown in the figure, only 6 of the 58 silver carp samples were misjudged, and the overall accuracy of the interactive validation was as high as 89.66%.
[0141] This suggests that a logistic regression model constructed using alanine, histidine, and arginine as biomarkers and their relative percentages can be used to differentiate wild and farmed silver carp from the Yangtze River basin. The excellent receiver operating characteristic (ROC) curve results, high correct diagnostic rates, and good biomarker complementarity suggest that these three amino acids have the potential to distinguish wild from farmed silver carp.
[0142] Amino acid biomarker composition for distinguishing wild silver carp from farmed silver carp and its use
[0143] 1. Selection of amino acid biomarker compositions
[0144] This example tests the amino acid content in muscle samples of wild and farmed silver carp, and trains a binary logistic regression model based on the test results. Amino acids with high statistical significance, low measurement error, and high data integrity are selected as biomarkers for distinguishing wild from farmed silver carp. The amino acid biomarkers provided in this example include alanine and arginine.
[0145] The embodiments also provide a kit for detecting these biomarkers, which includes reagents for detecting these biomarkers, such as reagents required for automatic amino acid analyzer testing.
[0146] 2. How to distinguish wild silver carp from farmed silver carp
[0147] The embodiments also disclose a method for distinguishing wild silver carp from farmed silver carp. The method includes: obtaining a fourth logistic regression model and a fourth threshold; obtaining multiple relative content percentages of multiple biomarker levels in a test sample; inputting the multiple relative content percentages into the fourth logistic regression model to obtain a fourth wild probability; and determining whether the test sample is wild or farmed silver carp based on the difference between the fourth wild probability and the fourth threshold. The multiple biomarkers include alanine and arginine.
[0148] In some embodiments, the step of obtaining the fourth logistic regression model includes:
[0149] 1) Obtain training samples
[0150] Example: 30 wild silver carp muscle samples and 30 farmed silver carp muscle samples were tested using the same automatic amino acid analyzer described above to obtain the peak areas for alanine and arginine in each sample. The relative content percentages of the target amino acids were calculated using the respective calculation formulas. The relative content percentages of alanine and arginine in these muscle samples served as training samples.
[0151] 2) Training
[0152] A binary logistic regression model is trained using the relative percentages of alanine and arginine as independent variables and the probability of a sample being wild as the dependent variable. In some embodiments, the binary logistic regression model training process includes: performing a composition analysis on the dependent and independent variables to determine whether they meet the prerequisites for logistic regression; performing a significance test on the independent variables, including degrees of freedom, significance, and score; performing a Hosmer-Lemmeshow test to verify the applicability and rationality of the model and ensure its accuracy; and training the binary logistic regression model to obtain a fourth logistic regression model based on the degree of influence of the independent variables on the dependent variable.
[0153] In some steps, the embodiment performed a single degree of freedom test on the independent variables of alanine and arginine in muscle sample content, including providing degrees of freedom, significance, and score testing. The results are shown in Table 15. The overall statistical scores of these independent variables are high, with significance less than 0.001, indicating that the independent variables have a very significant impact on the corresponding variables overall.
[0154] Table 15 Significance test and score of independent variables in the joint biomarker model of wild and farmed silver carp
[0155]
[0156] In some steps, the examples performed a Hosmer-Lemmeshow test on the logistic regression models for biomarkers in wild and farmed silver carp. Under multiple degrees of freedom, the significance of the binary logistic regression model was P = 0.633 > 0.05, indicating that the null hypothesis was accepted. As shown in Table 16, the constructed binary logistic regression model fit the real data well and reliably reflected the reliable relationships between the original variables.
[0157] Table 16 Hosmer-Lemmeshow test for binary logistic regression model of wild and farmed silver carp
[0158] step Chi-square degrees of freedom Significance 1 6.130 8 0.633
[0159] 3) Obtain the fourth logistic regression model
[0160] In some steps, the embodiment performs binary logistic regression analysis on the biomarkers for distinguishing wild silver carp from farmed silver carp. Table 17 shows the effect size Exp(B) of the regression model and the significance of each biomarker.
[0161] Table 17 Binary logistic regression analysis of wild and farmed silver carp
[0162]
[0163] Example According to the relative content percentages of the muscle samples of 30 wild silver carp and 30 farmed silver carp, a fourth logistic regression model was obtained as follows:
[0164] X=4.334×A+7.664×B-97.780;
[0165] The fourth wild probability = 1 / (1+e －X );
[0166] A and B are the relative content percentages of alanine and arginine, respectively, and the fourth wild probability represents the probability that the test sample is wild silver carp.
[0167] The steps provided in the embodiment are to obtain a fourth threshold value by ROC curve calculation. The steps provided in some embodiments include: using the relative content percentages of alanine and arginine in muscle samples of 30 wild silver carp and 30 farmed silver carp as a training set; training a binary logistic regression model with the training set to obtain multiple fourth wild probability values as described above; calculating the sensitivity (true positive rate) and 1-specificity (false positive rate) of each fourth wild probability value, and plotting the ROC curve with 1-specificity as the abscissa and sensitivity as the ordinate; calculating the Youden index corresponding to each predicted wild probability (Youden index = sensitivity + specificity - 1), and obtaining the maximum Youden index, wherein the fourth predicted wild probability corresponding to the maximum Youden index is the fourth threshold value.
[0168] In some embodiments, if the fourth wild probability is greater than a fourth threshold, the test sample is determined to be wild silver carp. If the fourth wild probability is less than the fourth threshold, the test sample is determined to be farmed silver carp.
[0169] In some embodiments, the fourth threshold value obtained by calculating the ROC curve using the above test set is 0.3020. That is, when the fourth wild probability is greater than 0.3020, the test sample is determined to be wild silver carp; when the fourth wild probability is less than 0.3020, the test sample is determined to be farmed silver carp.
[0170] 3. Identify application and testing
[0171] In some test cases, alanine and arginine from 58 muscle samples (including 28 wild and 30 farmed) were selected as biomarkers, and their relative content percentages were input into the above-mentioned fourth logistic regression model to obtain the fourth wild probability; based on the size of the fourth wild probability and the fourth threshold (0.3020), it was determined whether the test sample was wild silver carp or farmed silver carp.
[0172] Binary logistic regression models were constructed using the relative percentages of alanine and arginine in the training set, and a fourth logistic regression model was constructed using the combined relative percentages of alanine and arginine. These 58 muscle samples were tested, and the discrimination results were analyzed using a receiver operating characteristic (ROC) curve test, as shown in Table 18. As shown in Table 18, the fourth logistic regression model constructed using the combined relative percentages of alanine and arginine for discrimination of the test samples exhibited higher AUC values, sensitivity, and specificity.
[0173] Table 18 ROC test results of each biomarker and its combination for discriminating wild and farmed silver carp
[0174] Biomarkers AUC Sensitivity Specificity Alanine 0.850 1.000 0.667 Arginine 0.920 0.893 0.800 Composition 0.951 1.000 0.800
[0175] The discrimination results were input into SPSS 20.0 analysis software for confusion matrix test, such as Figure 4 As shown in the figure, only 6 of the 58 silver carp samples were misjudged, and the overall accuracy of the interactive validation was as high as 89.66%.
[0176] This suggests that using alanine and arginine as biomarkers and the regression model for discriminating wild and farmed silver carp is robust. The excellent ROC curve results, high correct diagnosis rates, and good biomarker complementarity suggest that these two amino acids have the potential to distinguish wild and farmed silver carp.
[0177] The above is only a preferred specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any changes or replacements that can be easily thought of by any technician familiar with this technical field within the technical scope disclosed in this application should be covered by the scope of protection of the present application.
Claims
1. Biomarkers for distinguishing wild silver carp from farmed silver carp include at least one of aspartic acid, alanine, phenylalanine, histidine, and arginine.
2. Methods for distinguishing wild silver carp from farmed silver carp include: Obtaining a first logistic regression model and a first threshold; Obtain multiple relative content percentages of multiple biomarkers in the test sample; Multiple biomarkers are aspartate, alanine, phenylalanine, histidine, and arginine; Inputting the plurality of relative content percentages into the first logistic regression model to obtain a first wild probability, wherein the first wild probability represents a probability that the test sample is wild silver carp; Determining whether the test sample is the wild silver carp or the farmed silver carp according to the difference between the first wild probability and the first threshold; Among them, the first logistic regression model is as follows: X=-2.707×A+2.776×B-1.281×C-2.114×D+5.323×E-29.477; First wild probability = 1 / (1+e －X ); Among them, A, B, C, D, and E are the relative content percentages of aspartic acid, alanine, phenylalanine, histidine, and arginine, respectively.
3. The method according to claim 2, wherein if the first wild probability is greater than the first threshold, the test sample is a wild silver carp; If the first wild probability is less than the first threshold, the test sample is farmed silver carp.
4. Methods for distinguishing wild silver carp from farmed silver carp include: Obtain a second logistic regression model and a second threshold; Obtain multiple relative content percentages of multiple biomarkers in the test sample; Multiple biomarkers are aspartate, alanine, histidine, and arginine; Inputting the plurality of relative content percentages into the second logistic regression model to obtain a second wild probability, wherein the second wild probability represents a probability that the test sample is wild silver carp; Determining whether the test sample is the wild silver carp or the farmed silver carp according to the second wild probability and the second threshold; Among them, the second logistic regression model is as follows: X=-2.769×A+2.938×B-2.378×C+5.239×D-34.888; Second wild probability = 1 / (1+e －X ); Among them, A, B, C, and D are the relative content percentages of aspartic acid, alanine, histidine, and arginine, respectively.
5. The method according to claim 4, wherein if the second wild probability is greater than the second threshold, the test sample is a wild silver carp; If the second wild probability is less than the second threshold, the test sample is farmed silver carp.
6. Methods for distinguishing wild silver carp from farmed silver carp include: Obtaining the third logistic regression model and the third threshold; Obtaining multiple relative content percentages of multiple biomarkers in the test sample, where the multiple biomarkers are alanine, histidine, and arginine; Inputting the plurality of relative content percentages into the third logistic regression model to obtain a third wild probability, wherein the third wild probability represents the probability that the test sample is wild silver carp; Determining whether the test sample is the wild silver carp or the farmed silver carp according to the third wild probability and the third threshold; Among them, the third logistic regression model is as follows: X=2.494×A-2.157×B+7.298×C-74.351; The third wild probability = 1 / (1+e －X ); Wherein A, B, and C are the relative content percentages of alanine, histidine, and arginine, respectively.
7. The method according to claim 6, wherein if the third wild probability is greater than the third threshold value, the test sample is a wild silver carp; If the third wild probability is less than the third threshold, the test sample is farmed silver carp.
8. Methods for distinguishing wild silver carp from farmed silver carp include: Obtaining a fourth logistic regression model and a fourth threshold; Obtaining multiple relative content percentages of multiple biomarkers in the test sample, wherein the multiple biomarkers are alanine and arginine; Inputting the plurality of relative content percentages into the fourth logistic regression model to obtain a fourth wild probability, wherein the fourth wild probability represents a probability that the test sample is wild silver carp; Determining whether the test sample is the wild silver carp or the farmed silver carp according to the fourth wild probability and the fourth threshold; Among them, the fourth logistic regression model is as follows: X=4.334×A+7.664×B-97.780; The fourth wild probability = 1 / (1+e －X ); Where A and B are the relative content percentages of alanine and arginine respectively.
9. The method according to claim 8, wherein if the fourth wild probability is greater than the fourth threshold value, the test sample is a wild silver carp; If the fourth wild probability is less than the fourth threshold value, the test sample is farmed silver carp.
10. Use of the biomarker according to claim 1 and / or a reagent for detecting any biomarker according to claim 1 in distinguishing wild from farmed silver carp.