A biomarker for diagnosing whether a subject is breast cancer

Abstract

The application provides a biomarker for breast cancer detection and application thereof, a proteomic method is used, proteins with significant differences in blood of breast cancer patients and normal persons are analyzed, a series of brand-new biomarkers capable of early indicating breast cancer occurrence risks are screened out, and a group of biomarkers are further screened out to construct a breast cancer diagnosis model, so that the breast cancer diagnosis model can be used for conveniently, non-invasively and efficiently predicting whether an individual suffers from breast cancer, and meets clinical requirements.

Description

[0001] This application is a divisional application of Chinese invention patent application No. 202310043352.4, filed on January 29, 2023. Technical Field

[0002] This invention relates to the medical field, and more specifically, to the use of proteomics to screen for biomarkers of breast cancer and to their use in the diagnosis of breast cancer, particularly to a biomarker for predicting the risk of developing breast cancer and its application. Background Technology

[0003] Proteomics is the science that studies the composition, location, changes, and interactions of proteins in cells, tissues, or organisms, including the study of protein expression patterns and proteome functional patterns. With the development of mass spectrometry, liquid chromatography-mass spectrometry (LC-MS / MS) has become the most important tool in proteomics research. The advancements in proteomics are significant for identifying diagnostic markers for diseases, screening drug targets, and conducting toxicological studies, and are therefore widely used in medical research.

[0004] Breast cancer is the most common cancer among women worldwide and the leading cause of cancer death in women. According to the latest global cancer burden data released by the International Agency for Research on Cancer (IARC) of the World Health Organization in 2020, there were 2.26 million new cases of breast cancer worldwide in 2020, surpassing lung cancer (2.21 million cases) for the first time to become the leading cause of cancer death globally.

[0005] Mounting evidence suggests that various factors, including genetic and environmental factors, may be involved in the occurrence and development of breast cancer. Early diagnosis of breast cancer is a crucial aspect of its treatment. Among various diagnostic platforms, imaging techniques are the primary diagnostic tool, providing valuable data for breast cancer patients. Furthermore, biochemical biomarkers, such as proteins, DNA, mRNA, and microRNA, can serve as novel diagnostic and therapeutic tools for breast cancer. With advancements in immunology and molecular biology, tumor-associated protein biomarkers are demonstrating increasingly important clinical value in the diagnosis and treatment of breast cancer, becoming indispensable biological indicators for auxiliary diagnosis, monitoring treatment efficacy, and predicting prognosis.

[0006] Clinically, a variety of tumor markers have been discovered that can be used for the diagnosis, pathological classification, clinical staging, prognosis, and efficacy assessment of breast cancer. However, the diagnostic efficacy of commonly used breast cancer markers (CA153, CA125, CEA) is not ideal, and no specific tumor marker has been found that has high sensitivity and specificity for the diagnosis of breast cancer.

[0007] Therefore, finding new biomarkers for breast cancer diagnosis, combining multiple biomarkers, and using appropriate breast cancer diagnostic prediction models have significant clinical value. Summary of the Invention

[0008] To address the problems existing in the prior art, this invention provides a biomarker for breast cancer detection. Utilizing proteomics methods, it analyzes proteins with significant differences in the blood of breast cancer patients and healthy individuals to screen out a series of novel biomarkers that can predict the early risk of breast cancer. From these, a further set of biomarkers is selected to construct a diagnostic model for breast cancer, which can be used to conveniently, non-invasively, and efficiently predict whether an individual has breast cancer, meeting clinical needs.

[0009] On one hand, the present invention provides the use of a biomarker in the preparation of a reagent for predicting whether an individual has breast cancer, said biomarker being selected from one or more of the following or composed of the following biomarkers: DCD, PLIN1, FABP4, SFRP1, S100B, CRYAB, PGM5, CD36 and ADH1B.

[0010] This invention utilizes TMT-labeled quantitative proteomics research, employing LC-MS / MS ultra-high performance liquid chromatography-tandem mass spectrometry to analyze blood samples from two groups: a healthy group and a breast cancer patient group. Then, orthogonal partial least squares analysis is used to identify proteins with significant differences between the breast cancer samples and control samples. Ultimately, nine proteins associated with breast cancer are identified as biomarkers that can be used to efficiently and accurately predict whether an individual has breast cancer.

[0011] In some approaches, the biomarkers that can be used to predict whether an individual has breast cancer can be used to prepare detection reagents, such as sample pretreatment reagents, antigens or antibodies, and other biological reagents and kits suitable for the detection of the biomarkers; or they can be developed into standardized reagents or kits suitable for the detection of the biomarkers by LC-UV or LC-MS.

[0012] In some embodiments, the markers are as follows: dermatophytein (DCD) is the protein or amino acid sequence of UniProt database number P81605; lipolysin-1 (PLIN1) is the protein or amino acid sequence of UniProt database number O60240; fatty acid-binding protein, adipocyte (FABP4) is the protein or amino acid sequence of UniProt database number P15090; secreted curl-associated protein 1 (SFRP1) is the protein or amino acid sequence of UniProt database number Q8N474; S100 calcium-binding protein B (S100B) is the protein or amino acid sequence of UniProt database number P04271; α-lens protein B. CRYAB is the protein or amino acid sequence with UniProt database number P02511; PGM5 is the protein or amino acid sequence with UniProt database number Q15124; CD36 is the protein or amino acid sequence with UniProt database number P16671; and all-trans retinol dehydrogenase [NAD(+)] (ADH1B) is the protein or amino acid sequence with UniProt database number P00325.

[0013] Furthermore, the biomarkers include DCD, PLIN1, FABP4, SFRP1, S100B, CRYAB, PGM5, CD36, and ADH1B.

[0014] In some embodiments, the biomarkers include dermatophytein (DCD); lipolysin-1 (PLIN1); fatty acid-binding protein, adipocyte (FABP4); secreted coil-associated protein 1 (SFRP1); S100 calcium-binding protein B (S100B); α-lens protein B chain (CRYAB); phosphoglucose mutase-like protein 5 (PGM5); platelet glycoprotein 4 (CD36); and all-trans retinol dehydrogenase [NAD(+)] (ADH1B).

[0015] Furthermore, the reagent is used to detect biomarkers in body fluid samples, which include any one of blood, urine, saliva, and sweat.

[0016] In some embodiments, the biomarkers of the present invention are obtained through screening blood samples, and are particularly suitable for development into blood test reagents or kits for breast cancer prediction.

[0017] This invention identifies biomarkers for breast cancer by screening blood samples. These biomarkers differ significantly in the blood of breast cancer patients and non-breast cancer patients. By collecting blood samples, these biomarkers in an individual's blood can be detected to predict or assist in the diagnosis of whether an individual has breast cancer or the likelihood of having breast cancer. Alternatively, these biomarkers in the blood of a group can be detected, thereby dividing the group into a breast cancer group or a non-breast cancer group.

[0018] Furthermore, the biomarkers detected in the body fluid sample refer to the presence, relative abundance, or concentration of biomarkers in the body fluid sample of the individual.

[0019] In some methods, relative abundance is preferred, which is the peak area of ​​the biomarker in the detection chromatogram obtained by high performance liquid chromatography-tandem mass spectrometry. For example, if the average peak area of ​​a biomarker measured in a control sample (an individual without breast cancer) is 500, and the average peak area measured in a breast cancer sample is 3000, then the abundance of the biomarker in the breast cancer sample is considered to be 6 times that in the control sample.

[0020] On the other hand, the present invention provides a combination of biomarkers for predicting whether an individual has breast cancer, the biomarkers being selected from the following biomarkers: DCD, PLIN1, FABP4, SFRP1, S100B, CRYAB, PGM5, CD36, and ADH1B.

[0021] Data from clinical breast cancer samples showed that using only these nine biomarkers to predict breast cancer achieved an AUC value of 0.964, which is significantly better than existing models that combine biomarkers to predict breast cancer.

[0022] On the other hand, the present invention provides a kit for predicting whether an individual has breast cancer, the kit comprising detection reagents as described above for detecting the above-mentioned 9 biomarkers, or combinations of biomarkers as described above.

[0023] In some embodiments, the detection reagent is an antibody against the biomarker described above, wherein the antibody is a monoclonal antibody.

[0024] In another aspect, the present invention provides a system for predicting whether a subject has breast cancer. The system includes a data analysis module for analyzing the detection values ​​of biomarkers in the subject. The biomarkers are a combination of the following biomarkers: DCD, PLIN1, FABP4, SFRP1, S100B, CRYAB, PGM5, CD36, and ADH1B; or.

[0025] Furthermore, the data analysis module calculates a predictive value for whether an individual has breast cancer by substituting the detected values ​​of the subject's biomarkers into the following equation, thereby assessing whether the individual has breast cancer. The equation is:

[0026] Where y is the predicted value, i represents the i-th biomarker, and n represents the number of biomarkers, where n=9, x i The value of the i-th biomarker is expressed in μg / mL; c represents the prediction type: negative or positive, P(y) c ) represents the prior probability of predicting the type c, where the prior probabilities of both negative and positive samples are 0.5; σ ci and μ ci The distribution represents the marker x when the category y is c. i The corresponding standard deviation and mean, the stated x i The corresponding standard deviations and means are shown in the table below: .

[0027] Furthermore, when the diagnostic model predicts a value Y > 0.513, the subject is considered to be a breast cancer patient; when the model predicts a value Y ≤ 0.513, the subject is considered not to be a breast cancer patient.

[0028] Furthermore, it also includes a data detection system and a data input / output interface; the data detection system is used to detect biomarkers in the sample and obtain detection values; the input interface of the data input / output interface is used to input the detection values ​​of biomarkers, and after the data analysis module analyzes the detection values, the output interface is used to output the analysis results of whether the individual has breast cancer.

[0029] In another aspect, the present invention provides the use of the system described above for constructing a detection model that predicts the probability of whether an individual has breast cancer.

[0030] The beneficial effects of this invention are as follows: Nine novel biomarkers were identified that can predict the risk of breast cancer at an early stage: DCD, PLIN1, FABP4, SFRP1, S100B, CRYAB, PGM5, CD36, and ADH1B. Different diagnostic models for breast cancer were constructed using nine biomarkers. The best model was found to be constructed using nine biomarkers, including DCD, PLIN1, FABP4, SFRP1, S100B, CRYAB, PGM5, CD36, and ADH1B. This model can be used to more efficiently predict whether an individual will have breast cancer, with an AUC value of 0.984, which is significantly better than existing breast cancer diagnostic models.

[0031] For the same marker substance, different models achieve different diagnostic results, and the most effective diagnostic prediction model is selected. Attached Figure Description

[0032] Figure 1 The image shows the Wilcoxon results from the TCGA database for the healthy control and breast cancer groups in Example 1.

[0033] Figure 2 The image shows the Wilcoxon results from the CPTAC database for the healthy control and breast cancer groups in Example 1.

[0034] Figure 3 This is a Venn diagram of differentially expressed genes and blood proteins from the TCGA and CPTAC databases in Example 1.

[0035] Figure 4 The graph shows the ROC and Boruta analysis results for the healthy control and breast cancer groups in Example 1.

[0036] Figure 5 This is a graph showing the performance evaluation results of the optimal model constructed based on nine different algorithms in Example 3.

[0037] Figure 6 The ROC curve of the breast cancer combined diagnostic model constructed in Example 3 is shown in the model group.

[0038] Figure 7 The ROC curve of the breast cancer combined diagnostic model constructed in Example 3 is shown in the test group.

[0039] Figure 8 This is a graph showing the performance evaluation results of the breast cancer combined diagnostic model constructed in Example 3 in the test group.

[0040] Detailed description (1) Diagnosis or testing Diagnosis or testing here refers to the detection or analysis of biomarkers in a sample, or the determination of the content of a target biomarker, such as its absolute or relative content. The presence or quantity of the target biomarker then indicates whether the individual providing the sample may have or suffer from a certain disease, or the likelihood of having a certain disease. The meanings of diagnosis and testing are interchangeable here. The results of such testing or diagnosis cannot be directly considered as a direct result of disease; rather, they are intermediate results. To obtain a direct result, further auxiliary methods such as pathology or anatomy are needed to confirm the presence of a specific disease. For example, this invention provides several novel biomarkers associated with breast cancer, and changes in the levels of these biomarkers are directly related to the presence or absence of breast cancer.

[0041] (2) The link between biomarkers or markers and breast cancer In this invention, "marker" and "biomarker" have the same meaning. Here, "linkage" refers to a direct correlation between the presence or change in the level of a certain biomarker in a sample and a specific disease; for example, a relative increase or decrease in the level indicates a higher likelihood of having the disease compared to healthy individuals.

[0042] If multiple different biomarkers are present simultaneously in a sample, or if their relative levels change, it indicates a higher likelihood of having the disease compared to healthy individuals. In other words, among biomarkers, some are strongly associated with the disease, some are weakly associated, and some may not even be associated with a specific disease. One or more of the strongly associated biomarkers can be used as diagnostic markers, while weakly associated biomarkers can be combined with strong biomarkers to diagnose a disease, increasing the accuracy of test results.

[0043] The numerous serum biomarkers discovered in this invention can be used to differentiate between breast cancer and healthy individuals. These biomarkers can be used individually for direct detection or diagnosis, indicating a strong correlation between the relative change in their levels and breast cancer. Of course, it is understood that one or more biomarkers strongly associated with breast cancer can be detected simultaneously. It is generally understood that in some methods, selecting highly correlated biomarkers for detection or diagnosis can achieve a certain level of accuracy, such as 60%, 65%, 70%, 80%, 85%, 90%, or 95%. This indicates that these biomarkers can provide intermediate values ​​for diagnosing a certain disease, but does not necessarily mean a direct diagnosis of that disease.

[0044] Of course, differentially expressed proteins with higher ROC values ​​can also be selected as diagnostic biomarkers. The terms "strong" and "weak" are generally determined using algorithms, such as biomarker contribution rate to breast cancer or weighted analysis. Such calculation methods can include significance analysis (…). p The values ​​(or FDR values) and fold changes are used in multivariate statistical analysis, which mainly includes principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA), and orthogonal partial least squares discriminant analysis (OPLS-DA). Other methods, such as ROC analysis, are also included. Of course, other model prediction methods are also possible. When specifically selecting biomarkers, the differentially expressed proteins disclosed in this invention can be chosen, or combinations of other existing known biomarkers can be selected and predicted using model methods. Detailed Implementation

[0045] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be noted that the embodiments described below are intended to facilitate understanding of the present invention and are not intended to limit it in any way. The reagents used in this embodiment are all known products and were obtained by purchasing commercially available products.

[0046] Example 1: Screening for biomarkers of breast cancer using public databases Data collection RNA sequencing data of breast cancer tissue samples were downloaded from the TCGA database (https: / / portal.gdc.cancer.gov / ). The tissue sample selection criteria were: pathologically confirmed primary breast cancer with complete clinical data and follow-up results. A total of 1097 primary breast cancer tissue samples and 114 normal tissue (healthy) samples were obtained for subsequent analysis. Proteomics data of breast cancer tissue samples, including 143 primary breast cancer tissue samples and 114 normal tissue (healthy) samples, were downloaded from the CPTAC database (https: / / proteomics.cancer.gov / programs / cptac) for subsequent analysis.

[0047] Data preprocessing RNA sequencing data downloaded from the TCGA database was first normalized using the TPM method to extract the raw read count data. Then, the normalized TPM values ​​were logarithmically transformed using Log2 to ensure a normal distribution. To obtain high-quality quantification results, genes (RNAs) with an average expression value less than 6 across all samples were removed. This resulted in RNA quantification results for 12,427 genes. Proteomics data downloaded from the CPTAC database was first screened, and relative quantification ratios were calculated using specific peptides. These ratios were then logarithmically transformed using Log2 to ensure a normal distribution. To obtain high-quality quantification results, proteins with quantification information in fewer than 10 samples were removed. This resulted in RNA quantification results for 7,340 proteins.

[0048] Difference analysis A combination of univariate and multivariate statistical analyses was used to screen for differentially expressed proteins and transcripts. Univariate analysis primarily included significance analysis (p-value or FDR value) and fold change of characteristic molecules in different groups. Multivariate statistical analyses mainly included receiver operating characteristic (ROC) curve analysis and Boruta feature screening based on the random forest algorithm. All statistical analyses were performed using R; specific R-related information is shown in Table 2.

[0049] Table 2: R used in this invention and related information

[0050] Variable Importance for the Projection (VIP) was calculated to measure the influence and explanatory power of each protein's expression pattern on the classification of each group of samples. A Wilcoxon rank-sum test was then performed to obtain the corrected VIP. p FDR (Fold Change Value). Based on the criteria of FDR < 1e-10 and Fold change > 2, 104 downregulated proteins and 8 upregulated proteins were obtained from CPTAC-derived proteomics data (see details). Figure 1 ), 359 downregulated and 232 upregulated genes were obtained from RNA sequencing data from TCGA (see details). Figure 2 ).

[0051] To more accurately identify potential breast cancer biomarkers, we compared differentially expressed genes screened from the CPTAC and TCGA databases with blood proteins from The Human Protein Atlas database. We defined 23 genes from the intersection of these three databases as potential breast cancer biomarkers (see details). Figure 3 ).

[0052] To evaluate the role of each biomarker in breast cancer diagnosis prediction, we used ROC and Boruta analyses to evaluate each biomarker. The results are shown in the figure below. Figure 4 The x-axis represents the AUC obtained from ROC analysis, the y-axis represents -log10(FDR) calculated by the Wilcoxon test, and the size of the point represents the VIP value obtained from the Boruta analysis.

[0053] Among the 23 candidate biomarkers, further screening was conducted based on VIP>5 and AUC>0.9, resulting in 9 more significant candidate biomarkers, as detailed in Table 3.

[0054] Table 3: Differential biomarkers between breast cancer and normal health

[0055] In Table 3, a smaller FDR value and / or a larger VIP value indicate, to some extent, that the difference between the two groups of the differential compound is more significant, and also suggest that the differential compound may have higher diagnostic value.

[0056] The newly discovered differential biomarkers for breast cancer can serve as candidate biomarkers for the differential diagnosis of breast cancer and healthy individuals. Selecting one or more of these biomarkers in combination can be used for the auxiliary diagnosis of breast cancer or to differentiate between breast cancer patients and healthy individuals.

[0057] Example 2: Nine single biomarkers predict breast cancer This embodiment uses a single biomarker selected in Example 1 to establish a predictive or diagnostic model for breast cancer, which can be used to distinguish between breast cancer and non-breast cancer, or to screen for breast cancer patients from a population, or to predict whether an individual is a breast cancer patient or the likelihood of an individual or test subject developing breast cancer.

[0058] ROC curves were established for each of the nine proteins provided in Example 1, and the area under the curve (AUC) was used to determine the quality of the experimental results. An AUC of 0.5 indicates that the individual protein has no diagnostic value; an AUC greater than 0.5 indicates that the individual protein has diagnostic value; the larger the AUC, the higher the diagnostic value of the individual protein. The results are shown in Table 4.

[0059] Table 4: ROC values ​​and related information of differentially expressed proteins between breast cancer and normal healthy samples based on ROC analysis.

[0060] The correlation between changes in the concentration of the nine biomarkers and the presence or absence of breast cancer can be distinguished using AUC values, sensitivity, and specificity, as shown in Table 4. Among these, the AUC value is the most intuitive and obvious. A higher AUC value indicates that the biomarker can more accurately distinguish between breast cancer patients and non-breast cancer patients.

[0061] As shown in Table 4, the concentration changes of the nine biomarkers are significantly associated with breast cancer. When any one of the nine biomarkers is used alone to distinguish between breast cancer patients and non-breast cancer patients, the AUC value can reach above 0.7, indicating high accuracy. Among them, S100 calcium-binding protein B (S100B) has the highest correlation, with an AUC value of 0.768.

[0062] Example 3: A classification model for identifying breast cancer patients and healthy individuals using nine differentially expressed proteins and its establishment. While a single biomarker can distinguish between breast cancer and non-breast cancer serum samples or predict breast cancer, generally combining multiple biomarkers results in higher accuracy in differentiation or prediction.

[0063] However, a single biomarker that is more accurate in predicting breast cancer does not necessarily play a greater role in the combination with one or more other biomarkers. Also, the more biomarkers there are, the higher the predictive accuracy (AUC value) of the combination is not necessarily true. Therefore, a large number of validation experiments are still needed.

[0064] This embodiment studies a model constructed from nine protein markers: serum dermatophytein (DCD); lipolysin-1 (PLIN1); fatty acid-binding protein adipocyte (FABP4); secreted coil-associated protein 1 (SFRP1); S100 calcium-binding protein B (S100B); α-lens protein B chain (CRYAB); phosphoglucose mutase-like protein 5 (PGM5); platelet glycoprotein 4 (CD36); and all-trans retinol dehydrogenase [NAD(+)] (ADH1B).

[0065] Get data Study population: From August to December 2019, 500 breast cancer patients and 500 healthy controls were collected. All enrolled patients signed informed consent forms. Breast cancer patients were all pathologically confirmed by biopsy, while healthy controls were individuals with normal routine physical examinations (no nodules or breast cancer). The enrolled participants were divided into a model group (breast cancer n=400, healthy controls n=400) and a test group (breast cancer n=100, healthy controls n=100) at a ratio of 8:2. Data are shown in Table 5. Table 5: Modeling Sample Information

[0066] Inclusion criteria for breast cancer patients: (a) no history of other malignant tumors, and (b) surgical treatment within one month of blood collection, with postoperative pathological confirmation of breast cancer. Healthy individuals in the control group were selected from a health checkup center; these individuals were confirmed to have no breast nodules and no history of malignant tumors by chest X-ray or thin-section computed tomography. All collected serum samples were stored in a serum bank at -80°C after obtaining informed consent.

[0067] In this embodiment, the collected serum samples were subjected to enzyme-linked immunosorbent assay (ELISA) to obtain the concentrations of nine protein markers in serum: dermatoglycin (DCD); lipolysin-1 (PLIN1); fatty acid-binding protein adipocyte (FABP4); secreted coil-associated protein 1 (SFRP1); S100 calcium-binding protein B (S100B); α-lens protein B chain (CRYAB); phosphoglucose mutase-like protein 5 (PGM5); platelet glycoprotein 4 (CD36); and all-trans retinol dehydrogenase [NAD(+)] (ADH1B), in μg / mL.

[0068] Experimental data statistical analysis The Shapiro Wilk test was used to assess normality, and the nonparametric Wilcoxon test was used to analyze differences in blood biomarker concentrations between breast cancer patients and healthy controls in both the model and test groups. In the model group, a combined diagnostic model for eight breast cancer biomarkers was constructed using a combination of machine learning methods. The area under the receiver operator characteristic (ROC) curve (AUC) was estimated using predicted probability values ​​with a 95% confidence interval (CI) to assess the discriminative power of the multivariate diagnostic model. In the test group, the Youden index (YI) was calculated to determine the predicted probability cut-off value used to distinguish breast cancer patients from normal controls. Furthermore, ROCs for individual biomarkers and different subgroups were constructed and compared. Standard descriptive statistics, such as frequency, mean, median, positive predictive value (PPV), negative predictive value (NPV), and standard deviation (SD), were calculated to describe the experimental results in the study population. Statistical analysis was performed using R 3.6.1, and a p-value less than 0.05 was considered statistically significant.

[0069] Steps for constructing a combined diagnostic model for breast cancer (9MP) S101 uses the concentration matrix of nine protein biomarkers from the model group samples as the original training dataset: dermal ...

[0070] S102 sets the supervised classification algorithm used to build the prediction model, and the grid search range during the hyperparameter optimization process. The supervised classification algorithms include eight algorithms: neural networks, gradient boosting, generalized linear models, random forests, logistic regression, support vector machines, Naive Bayes, and mixture discriminant analysis. In this step, the grid search range for hyperparameter optimization of the model is set for each algorithm, as shown in Table 6 below.

[0071] Table 6: Parameter grid search range of 8 algorithms

[0072] S103. Based on the algorithm and hyperparameter setting range set in step S102, select one algorithm and the corresponding hyperparameter combination method as the parameters for constructing the prediction model.

[0073] S104. The original dataset is split into K subsets using a K-fold cross-validation mechanism. To ensure that the proportion of majority and minority class samples in each subset is the same as in the original dataset, a stratified K-fold cross-validation mechanism is used for data splitting.

[0074] S105. Based on the K training data subsets obtained from step S104, select one subset as the validation set Ddev.

[0075] S106, combine the unselected subset of training data from step S105 to form the training data pool Dtrainl.

[0076] S107. Based on the training dataset D.train obtained in step S106, construct a prediction model based on the selected supervised classification algorithm and hyperparameters.

[0077] S108: Based on the prediction model obtained in step S107, evaluate it on the validation set D.dev to obtain the AUC value, and store the current prognostic prediction model and the corresponding AUC value in the prediction model pool. Step S108 involves evaluating the prediction model obtained in step S107 on the validation set determined in the current iteration, and storing both the model and the evaluation result in the prediction model pool for future use. The evaluation mentioned in this step can be the AUC value or other reasonable metrics for evaluating model performance.

[0078] S109, determine if each subset has been used as a validation set. Step S109 checks if all K subsets obtained in step S104 have been used as validation sets for model training. If all subsets have been used as validation sets and training has been completed, proceed to step S110; otherwise, proceed to step S105. This step ensures that every sample in the original dataset has been used as a validation set, improving model stability and preventing the model from overfitting to a particular subset.

[0079] S110: The average AUC of all models in the prediction model pool (Pool) is used as the final performance evaluation value of the model for this combination method. The model parameters and the final performance evaluation AUC value are then stored in the optimal model pool (Pool.best).

[0080] S111: Determine whether all algorithms and their corresponding hyperparameter combinations have been used to construct a prediction model. Step S111 involves determining whether all algorithms and their corresponding hyperparameter combinations obtained in step S102 have been used to construct a prediction model. If all combinations have completed model construction, proceed to step S112; otherwise, if any combination has not completed model construction, proceed to step S103.

[0081] S112, from the optimal model pool Pool.best obtained after the iteration of step S111, select the prediction model with the highest AUC value for each algorithm and store it in the candidate prediction model set M.set for breast cancer diagnosis.

[0082] S113, The model set M.set obtained in step S112 is evaluated in the test set D.test to obtain the AUC value. The model with the largest AUC value is selected as the final predictive model for breast cancer diagnosis. Parameter optimization results of the breast cancer combined diagnostic model (9MP) Following the model building steps outlined above, we obtained the optimal models for nine different algorithms. A 10-fold cross-validation method was used during the modeling process, and the model performance was evaluated based on AUC, sensitivity, and specificity. See Table 7 and... Figure 5 As shown, the Naive Bayes algorithm achieved the highest performance scores in AUC, sensitivity, and specificity among the three methods (AUC: 0.977; sensitivity: 0.905; specificity: 0.925).

[0083] Table 7: AUC of the model built under different hyperparameter combinations of the glmnet algorithm

[0084] Based on the above analysis results, the optimal model constructed by the Naive Bayes algorithm was selected as the final prediction model for breast cancer diagnosis. The equation for constructing this model is as follows:

[0085] Where y is the predicted value, i represents the i-th biomarker, n represents the number of biomarkers (n=9), and x i P(y) represents the detection value (μg / mL) of the i-th biomarker; c represents the prediction type (negative, positive); P(y) represents the detection value of the i-th biomarker. c ) represents the prior probability of predicting type c, where the prior probabilities of negative and positive samples are both 0.5; σ ci and μ ci The distribution represents the marker x when the category y is c. i The corresponding standard deviations and mean values ​​are detailed in Table 8 below.

[0086] Table 8: Mean and standard deviation of the concentrations of the nine biomarkers in the model across different sample types.

[0087] As can be seen from the above, for the same combination of biomarkers, different algorithms can lead to different models with significantly different diagnostic or predictive results. Therefore, when the biomarkers are the same, different algorithms are also crucial.

[0088] Determination of diagnostic thresholds for the combined diagnostic model for breast cancer (9MP) ROC curves were plotted using the predicted values ​​from the model group, and the optimal diagnostic cutoff value was set to 0.513 based on the Youden index. That is, when the model's predicted value was ≤0.513, the subject was considered not to have breast cancer; when the model's predicted value was >0.513, the subject was considered to have breast cancer. The results are as follows... Figure 6 As shown, the model's AUC in the model group was 0.978, its sensitivity was 90.5%, and its specificity was 93.5%.

[0089] Validation of the 9MP Combined Diagnostic Model for Breast Cancer ROC curves were plotted using the predicted values ​​from the test group, as shown in Figure 7, with an AUC of 0.984. The optimal diagnostic cutoff value was set to 0.513 based on the Youden index. That is, when the diagnostic model's predicted value was ≤0.513, the subject was considered neither a breast cancer patient nor a healthy individual; when the model's predicted value was >0.513, the subject was considered a breast cancer patient. The results are as follows... Figure 8As shown, the model achieved an accuracy of 93.5% in the test group, a Kappa value of 0.87, a sensitivity of 91.4%, a specificity of 95.8%, a positive predictive value of 96%, and a negative predictive value of 91%.

[0090] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the invention; therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. The use of biomarkers in the preparation of reagents for predicting whether an individual has breast cancer, characterized in that, The biomarkers consist of the following biomarkers: DCD, PLIN1, FABP4, SFRP1, S100B, CRYAB, PGM5, CD36, and ADH1B; the reagents are used to detect biomarkers in body fluid samples, which include any one of blood, urine, saliva, and sweat.

2. The use as described in claim 1, characterized in that, The detection methods include liquid chromatography-mass spectrometry (LC-MS).

3. The use as described in claim 1, characterized in that, The detection of biomarkers in body fluid samples includes the presence or concentration of biomarkers in the body fluid samples of the individual being tested.

4. As described in claim 1, wherein the DCD is a protein or amino acid sequence with UniProt database number P81605; PLIN1 is a protein or amino acid sequence with UniProt database number O60240; FABP4 is a protein or amino acid sequence with UniProt database number P15090; SFRP1 is a protein or amino acid sequence with UniProt database number Q8N474; S100B is a protein or amino acid sequence with UniProt database number P04271; CRYAB is a protein or amino acid sequence with UniProt database number P02511; PGM5 is a protein or amino acid sequence with UniProt database number Q15124; CD36 is a protein or amino acid sequence with UniProt database number P16671; and ADH1B is a protein or amino acid sequence with UniProt database number P00325.

Images