A detection method and application of benign and malignant breast tumors

Abstract

The present invention provides a method for detecting benign and malignant breast tumors and its application. Taking body fluid exosomes as samples, the present invention uses liquid chromatography tandem mass spectrometry to obtain the protein phenotypes of patients with benign and malignant breast tumors, establish a detection and / or prediction model for benign and malignant breast tumors based on proteomics, and provides a kit or chip for detecting benign and malignant tumors, as well as the application of protein markers for benign and malignant breast tumors in the preparation of the kit and / or chip. Compared with traditional detection methods such as tissue biopsy, the present invention has the characteristics of non-invasive and convenient detection, and can accurately distinguish whether the sample to be detected is a sample of benign or malignant breast tumor.

Description

Technical Field

[0001] The present invention relates to the technical fields of protein mass spectrometry detection and machine learning applied to cancer prediction and clinical medicine. Specifically, it relates to a method for detecting breast benign and malignant tumors, a kit or chip for detecting breast benign and malignant tumors, and the application of protein markers for breast benign and malignant tumors in the preparation of kits and / or chips. Background Art

[0002] Breast tumors are common diseases in general surgery, mostly occurring in young and middle-aged women, which can seriously affect the health status of patients. Clinically, breast tumors can be divided into two categories: benign and malignant. Their treatment plans and prognoses are completely different. If misdiagnosis leads to treatment failure, serious adverse consequences may occur. Therefore, the differential diagnosis of breast benign and malignant tumors is very important.

[0003] Currently, there are mainly five techniques for diagnosing breast benign and malignant tumors in China: (1) physical examination; (2) imaging diagnosis (mammography, breast ultrasound, breast magnetic resonance imaging); (3) tissue biopsy; (4) breast cancer tumor marker examination (serum cancer antigen 15-3 (CA15-3), serum carcinoembryonic antigen (CEA), serum cancer antigen 125 (CA125), etc.); (5) immunohistochemical examination (Ki-67, HER-2, ER, PR, etc.). However, there are the following deficiencies in the detection of breast cancer and breast benign tumors: (1) clinical physical examination is costly; (2) the positive miss detection rate of imaging diagnosis methods is high; (3) the currently relatively mature biological targets can only perform risk assessment and are not suitable for clinical diagnosis; (4) puncture biopsy is first invasive, and in addition, it has high requirements for the sampler and the puncture site, and it is easy to cause different results due to the influence of external factors. Therefore, there is an urgent need in this field to establish a new technique for examining breast benign and malignant tumors, which can sensitively and stably identify and diagnose breast benign and malignant tumors, so as to carry out timely and targeted treatment, improve the survival quality and survival rate of patients, reduce medical costs, and save social resources. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for detecting breast benign and malignant tumors, a kit or chip for detecting breast benign and malignant tumors, and the application of protein markers for breast benign and malignant tumors in the preparation of kits and / or chips, so as to provide a judgment on whether the sample to be tested is a breast benign or malignant tumor sample.

[0005] To achieve the above purpose, according to one aspect of the present invention, taking body fluid exosomes as samples, using liquid chromatography tandem mass spectrometry, the protein phenotypes of breast benign and malignant tumor patients are obtained, and a detection and / or prediction model for breast benign and malignant tumors is established.

[0006] Furthermore, the humoral specific markers of healthy control population, breast benign tumors, and breast malignant tumors (i.e., breast cancer) are used as a dataset to construct a detection and / or prediction model for breast benign and malignant tumors, and it is determined whether it is a breast benign / malignant tumor based on the output result of the model.

[0007] Furthermore, based on the humoral exosome samples of breast benign / malignant tumor patients and the humoral exosome samples of healthy control population, a training set and a validation set are established. The protein markers of the training set are classified into healthy control population, breast benign tumors, and breast malignant tumors (i.e., breast cancer), and a training detection and / or prediction model is constructed by using machine learning method.

[0008] Furthermore, for internal validation, the benign / malignant probabilities of the training set samples are output, an ROC curve is plotted, and the prediction result of the model is evaluated by the area under the ROC curve. For external validation, the benign / malignant probabilities of the validation set samples are output, an ROC curve is plotted, and the prediction result of the model is evaluated by the area under the ROC curve.

[0009] Furthermore, the humoral exosomes are blood, saliva, urine or cerebrospinal fluid exosomes of healthy people, breast benign patients, and breast malignant tumor (i.e., breast cancer) patients.

[0010] Preferably, the blood exosomes are serum exosomes or plasma exosomes.

[0011] Furthermore, the breast benign tumors are fibroadenomas or papillomas. The breast malignant tumors (i.e., breast cancer) are breast carcinoma in situ or invasive breast carcinoma according to the World Health Organization Classification of Tumours of the Breast; according to molecular typing, they are Luminal A type (ER and / or PR+, HER2-, Ki-67 < 20%), Luminal B type (HER2 negative type (type B1): ER+ and (or) PR < 20%, HER2-, Ki-67 ≥ 20%; B2: HER2 positive type (type B2): ER+ and (or) PR+, HER2 overexpression); HER2 positive type (HER2+, ER- and PR-) or basal-like / triple negative type (HER2-, ER- and PR-).

[0012] Furthermore, the application of any one or more of the following protein markers for breast benign / malignant tumors in the preparation of a kit and / or chip for detecting breast benign and malignant tumors: ACTR3, IGHV3-53, IGHV3-33, IGHV4-34, IGHV3-23, ARPC1A, IGHV4-59, AP2A1, SSC5D, ATP2B2, IGHV4-39, PUM1, LMAN2, COPE, ARF5, SEC31A, ADD2, PARK7, IQGAP1, FNBP1L, BZW1, PPL, RPL6, ADD1, PDLIM1, CNN2, SDCBP, CTNNA1, SNX1, TRIM29, BAG3, KIF5B, PCBP1, PRDX1, ERC1, SPTAN1, EIF2A, NOP56, VASP, PTPN1, STAT1, IDH1, RDX, GAPVD1, SND1, VASN, KTN1, EEF1G, TJP1, EHD1, AFDN, HNRNPK, CTTN, CAPZA1, RPL24, TAGLN2, MAPRE1, PFN1, RAN, DBN1, PLEC, EIF4G1, APP, PON1, APOA2, APOA1, PCSK9, APOA4, APOC2, LIPC, APOC1, ANGPTL3, APOF, APOE, APOB, APOL1, APOL2, F7, LDHA, HSP90AA1, IGFBP2, ANGPTL6, MMP2, PECAM1, CYP1B1, NAA15, TYMP, PGAM4, PGK2, HLA-H, HLA-B, HLA-C, HLA-G, CTSS, ITGB1, IGLV3-25, CD44, CXADR, IGLV3-27, PVR, and HCK.

[0013] Furthermore, the application of the combined use of 8 protein markers specific for breast benign / malignant tumors in the preparation of a kit and / or chip for detecting breast benign and malignant tumors: STAT1, PON1, APOC1, APOC2, MMP2, IGHV4-39, IGHV3-53, and ADD2.

[0014] Further, the kit and / or chip further includes a control. The expression level of any one or more of the following protein markers of breast benign tumors in exosomes of breast benign tumor patients is significantly higher than that of the control and the expression level in exosomes of breast malignant tumors, i.e., breast cancer patients, for detecting breast benign and malignant tumors: PARK7, IQGAP1, FNBP1L, BZW1, PPL, RPL6, ADD1, PDLIM1, CNN2, SDCBP, CTNNA1, SNX1, TRIM29, BAG3, KIF5B, PCBP1, PRDX1, ERC1, SPTAN1, EIF2A, NOP56, VASP, PTPN1, STAT1, IDH1, RDX, GAPVD1, SND1, VASN, KTN1, EEF1G, TJP1, EHD1, AFDN, HNRNPK, CTTN, CAPZA1, RPL24, TAGLN2, MAPRE1, PFN1, RAN, DBN1, PLEC, EIF4G1, APP, PON1, APOA2, APOA1, PCSK9, APOA4, APOC2, LIPC, APOC1, ANGPTL3, APOF, APOE, APOB, APOL1, APOL2, F7, LDHA, HSP90AA1, and IGFBP2.

[0015] Further, the kit and / or chip further includes a control. The expression level of any one or more of the following protein markers of breast malignant tumors, i.e., breast cancer, in exosomes of breast cancer patients is significantly higher than that of the control and the expression level in breast benign tumors, for detecting breast benign and malignant tumors: ANGPTL6, MMP2, PECAM1, CYP1B1, NAA15, TYMP, PGAM4, PGK2, HLA-H, HLA-B, HLA-C, HLA-G, CTSS, ITGB1, IGLV3-25, CD44, CXADR, IGLV3-27, PVR, and HCK.

[0016] Further, the kit and / or chip further includes a control. The expression level of any one or more of the following protein markers in exosomes of breast benign and malignant tumor patients is significantly lower than that of the control for detecting breast benign and malignant tumors: ACTR3, IGHV3-53, IGHV3-33, IGHV4-34, IGHV3-23, ARPC1A, IGHV4-59, AP2A1, SSC5D, ATP2B2, IGHV4-39, PUM1, LMAN2, COPE, ARF5, SEC31A, and ADD2.

[0017] Furthermore, the exosomes are serum exosomes from patients with benign breast tumors or serum exosomes from breast cancer patients.

[0018] Preferably, the breast malignant tumor, i.e., breast cancer, is ductal carcinoma in situ or invasive breast carcinoma according to the World Health Organization Classification of Tumors of the Breast.

[0019] Furthermore, the expression level is the expression level at the gene level, transcriptional level, or protein level.

[0020] Preferably, the expression level at the protein level is detected by one or more of mass spectrometry, Western blotting, and ELISA.

[0021] Furthermore, the breast malignant tumor, i.e., breast cancer, is Luminal A type (ER and / or PR+, HER2−, Ki-67 < 20%); Luminal B type (HER2 negative type (type B1): ER+ and / or PR < 20%, HER2−, Ki-67 ≥ 20%; B2: HER2 positive type (type B2): ER+ and / or PR+, HER2 overexpression); HER2 positive type (HER2+, ER−, and PR−), or basal-like / triple-negative type (HER2−, ER−, and PR−).

[0022] Furthermore, a kit for detecting breast benign and malignant tumors, the kit comprising reagents for identifying any one or more of the following breast benign / malignant tumor-specific protein markers: ACTR3, IGHV3-53, IGHV3-33, IGHV4-34, IGHV3-23, ARPC1A, IGHV4-59, AP2A1, SSC5D, ATP2B2, IGHV4-39, PUM1, LMAN2, COPE, ARF5, SEC31A, ADD2, PARK7, IQGAP1, FNBP1L, BZW1, PPL, RPL6, ADD1, PDLIM1, CNN2, SDCBP, CTNNA1, SNX1, TRIM29, BAG3, KIF5B, PCBP1, PRDX1, ERC1, SPTAN1, EIF2A, NOP56, VASP, PTPN1, STAT1, IDH1, RDX, GAPVD1, SND1, VASN, KTN1, EEF1G, TJP1, EHD1, AFDN, HNRNPK, CTTN, CAPZA1, RPL24, TAGLN2, MAPRE1, PFN1, RAN, DBN1, PLEC, EIF4G1, APP, PON1, APOA2, APOA1, PCSK9, APOA4, APOC2, LIPC, APOC1, ANGPTL3, APOF, APOE, APOB, APOL1, APOL2, F7, LDHA, HSP90AA1, IGFBP2, ANGPTL6, MMP2, PECAM1, CYP1B1, NAA15, TYMP, PGAM4, PGK2, HLA-H, HLA-B, HLA-C, HLA-G, CTSS, ITGB1, IGLV3-25, CD44, CXADR, IGLV3-27, PVR, and HCK.

[0023] Furthermore, a kit for detecting breast benign and malignant tumors, the kit comprising reagents for jointly identifying 8 breast benign / malignant tumor-specific protein markers: STAT1, PON1, APOC1, APOC2, MMP2, IGHV4-39, IGHV3-53, and ADD2.

[0024] Furthermore, the reagent is a mass spectrometry detection reagent or an antibody detection reagent for identifying the breast benign / malignant tumor-specific protein marker; preferably, the antibody detection reagent is a monoclonal antibody.

[0025] Furthermore, a chip for detecting benign and malignant breast tumors, wherein reagents for identifying any one or more of the following protein markers specific for benign / malignant breast tumors are provided on the chip: ACTR3, IGHV3-53, IGHV3-33, IGHV4-34, IGHV3-23, ARPC1A, IGHV4-59, AP2A1, SSC5D, ATP2B2, IGHV4-39, PUM1, LMAN2, COPE, ARF5, SEC31A, ADD2, PARK7, IQGAP1, FNBP1L, BZW1, PPL, RPL6, ADD1, PDLIM1, CNN2, SDCBP, CTNNA1, SNX1, TRIM29, BAG3, KIF5B, PCBP1, PRDX1, ERC1, SPTAN1, EIF2A, NOP56, VASP, PTPN1, STAT1, IDH1, RDX, GAPVD1, SND1, VASN, KTN1, EEF1G, TJP1, EHD1, AFDN, HNRNPK, CTTN, CAPZA1, RPL24, TAGLN2, MAPRE1, PFN1, RAN, DBN1, PLEC, EIF4G1, APP, PON1, APOA2, APOA1, PCSK9, APOA4, APOC2, LIPC, APOC1, ANGPTL3, APOF, APOE, APOB, APOL1, APOL2, F7, LDHA, HSP90AA1, IGFBP2, ANGPTL6, MMP2, PECAM1, CYP1B1, NAA15, TYMP, PGAM4, PGK2, HLA-H, HLA-B, HLA-C, HLA-G, CTSS, ITGB1, IGLV3-25, CD44, CXADR, IGLV3-27, PVR, and HCK.

[0026] Furthermore, a chip for detecting benign and malignant breast tumors, wherein reagents for jointly identifying 8 protein markers specific for benign / malignant breast tumors are provided on the chip: STAT1, PON1, APOC1, APOC2, MMP2, IGHV4-39, IGHV3-53, and ADD2.

[0027] Furthermore, the reagent is a mass spectrometry detection reagent or an antibody detection reagent for identifying the protein markers specific for the benign / malignant breast tumors; preferably, the antibody detection reagent is a monoclonal antibody.

[0028] Furthermore, the present invention also provides a system for detecting benign and malignant breast tumors, the system comprising:

[0029] 1. A biomarker combination verification module, which receives input data, judges the input data, and outputs a judgment result when the input data meets the judgment condition; the biomarker combination is the combination of all the above-mentioned 101 breast benign / malignant tumor-specific protein biomarkers. Preferably, the biomarker combination is STAT1, PON1, APOC1, APOC2, MMP2, IGHV4-39, IGHV3-53, and ADD2.

[0030] 2. A result judgment module, which predicts breast benign and malignant tumors according to the judgment result and outputs a prediction result.

[0031] Among them, in the biomarker combination verification module, when the input data meets judgment condition 1, the judgment result output is "Yes, benign"; when the input data meets judgment condition 2, the judgment result output is "Yes, malignant";

[0032] When the input data does not meet the judgment condition, the judgment result output is "No".

[0033] In the result judgment module, when the judgment result is "Yes, benign", the prediction result output is "Having breast benign tumor"; when the judgment result is "Yes, malignant", the prediction result output is "Having breast malignant tumor";

[0034] When the judgment result is "No", the prediction result output is "Not having breast benign and malignant tumors".

[0035] Furthermore, the system further includes a sample extraction module, which extracts sample data and transmits it to the biomarker combination verification module.

[0036] Furthermore, the present invention also provides an application of the above-mentioned system in the preparation of products for detecting breast benign and malignant tumors.

[0037] Furthermore, the present invention also provides a device including a machine learning model, which includes a memory and a processor. When the computer program stored in the memory is executed by the processor, the functions of the above-mentioned system can be realized.

[0038] Furthermore, the present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the functions of the above-mentioned system can be realized.

[0039] On the basis of conforming to the common knowledge in the art, the above-mentioned preferred conditions can be combined arbitrarily to obtain various preferred embodiments of the present invention.

[0040] The reagents and raw materials used in the present invention are all commercially available.

[0041] The positive and progressive effects of the present invention are as follows:

[0042] The present invention provides a method for detecting benign and malignant breast tumors and its applications. Using body fluid exosomes as samples, the present invention utilizes liquid chromatography-tandem mass spectrometry to obtain the protein phenotypes of patients with benign and malignant breast tumors and healthy control groups, establishes a detection model for benign and malignant breast tumors, and provides a kit or chip for detecting benign and malignant breast tumors, as well as the application of protein markers for benign and malignant breast tumors in the preparation of kits and / or chips. Compared with traditional detection methods such as tissue biopsy, the present invention has the characteristics of non-invasive and convenient detection, and can accurately determine whether the sample to be detected is a sample of benign or malignant breast tumor. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0044] Figure 1 It is a flowchart of a method for detecting benign and malignant breast tumors according to an embodiment of the present invention.

[0045] Figure 2 It is a cumulative curve graph of the proteome identification of serum exosomes according to an embodiment of the present invention.

[0046] Figure 3 It is a confusion matrix for internal verification of benign and malignant breast tumors. Healthy donors: HD; Benign mammary disease: BD; Breast cancer: BC.

[0047] Figure 4 It is a schematic diagram of the area under the ROC curve for internal verification.

[0048] Figure 5 It is a confusion matrix for external verification of benign and malignant breast tumors. Healthy donors: HD; Benign mammary disease: BD; Breast cancer: BC.

[0049] Figure 6 It is a schematic diagram of the area under the ROC curve for external verification. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0051] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments. However, it can be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0053] The present invention provides a method and application for predicting breast cancer metastasis. Taking body fluid exosomes as samples, the present invention uses liquid chromatography tandem mass spectrometry to obtain the protein phenotypes of breast benign and malignant tumor patients, establish a detection model for breast benign and malignant tumors, and provides a kit or chip for detecting breast benign and malignant tumors, and the application of breast benign and malignant tumor protein markers in the preparation of the kit and / or chip.

[0054] Figure 1 It is a flowchart of a method for detecting breast benign and malignant tumors according to an embodiment of the present invention. As Figure 1 shown, a method for detecting breast benign and malignant tumors according to an embodiment of the present invention includes the following steps:

[0055] (1) Selection of samples. Based on exosome samples derived from breast benign and malignant tumors, a training set and a validation set are established for proteomic analysis.

[0056] In one embodiment, the present invention selects serum exosome samples from 167 breast benign and malignant tumor patients and healthy control groups to establish a training set and a validation set for proteomic analysis.

[0057] Specifically, 70% of the samples are used as the training set, and 30% of the samples are used as the validation set. The training set includes 116 samples, and the validation set includes 51 samples.

[0058] (2) Sample preparation and mass spectrometry analysis. Exosome extraction, exosome protein extraction, mass spectrometry analysis, polypeptide identification and protein quantification are performed, and protein identification is completed through retrieval and comparison.

[0059] In one embodiment, a blood sample is drawn from each patient and serum separation is performed. First, serum exosomes are extracted. 1) Take 1 mL of serum and centrifuge it at 10,000 g for 30 minutes at 4°C. Dilute the supernatant with 20 mL of PBS and filter it through a 0.2-μm diameter filter membrane. 2) Transfer the filtered serum to an ultracentrifuge tube and centrifuge it at 150,000 g for 4 hours. Resuspend it with PBS and continue to centrifuge at 150,000 g for 2 hours at 4°C. 3) Discard the supernatant and retain approximately 200 μL of PBS.

[0060] Secondly, extraction and trypsin digestion of exosomal proteins. The exosome sample (usually 5 μg, adjusted according to BCA measurement) is dried by vacuum centrifugation and redissolved with 30 - 50 μL of 8 M urea / 50 mM ammonium bicarbonate / 10 mM DTT. After lysis and reduction, the protein is alkylated with 20 or 30 mM iodoacetamide (Sigma). The protein is digested overnight at 37°C with trypsin (Promega) at an enzyme-to-mass ratio of 1:50, and then the peptides are extracted and dried (SpeedVac, Eppendorf).

[0061] Thirdly, mass spectrometry analysis. We employed an EASY-nLC 1200 ultra-high pressure system liquid chromatography system (ThermoFisher Scientific). The peptides were separated within 75 min at a flow rate of 600 nL / min on a column with an I.D. of 150 μm × 15 cm, packed with 1.9-μm ReproSil-Pur 120 C18-AQ particles. Mobile phases A and B were water and acetonitrile (0.1 vol% FA), respectively. Within 75 min, %B increased linearly from 15 to 30%. The sample was analyzed on a Q-EX-HF mass spectrometer (MS) through a nanoelectrospray ion source (Thermo Fisher Scientific). For the generation of an ion mobility enhanced spectral library, the mass spectrometry was operated in data-independent mode. Generally, 75% of the sample was injected. The peptides were dissolved with 12 μL (0.1% formic acid) buffer, 9 μL of which was loaded onto a 100-μm I.D. × 2.5-cm C18 trapping column with a maximum pressure of 280 bar, and the solvent A (0.1% formic acid) was 14 μL. The DIA method included an MS1 scan from 300 - 1400 m / z at 60 k resolution (AGC target 4e5 or 50 ms). Then 30 DIA fragments were obtained at 15 k resolution with an AGC target of 5e4 and a maximum injection time of 22 ms. The "inject all available parallel times" setting was enabled. The HCD fragmentation was set to a normalized collision energy of 30%. The spectra were recorded in profile mode. The default charge state for MS2 was set to 3.

[0062] Finally, polypeptide identification and protein quantification. All data were processed using the in-house developed Firmiana one-stop proteomics cloud platform. The UniProt human protein database (updated on December 17, 2019, with 20,406 entries) was searched using FragPipe (v.12.1) and MSFragger (2.2). The mass tolerance for precursors was 20 ppm, and the mass tolerance for product ions was 50 mmu. Up to two missed cleavages were allowed. The search engine considered carbamidomethylation of cysteine as a fixed modification and N-acetylation and oxidation of methionine as variable modifications. The precursor ion score charge was restricted to +2, +3, and +4. The results of DIA data were combined into a spectral library using SpectraST software. All libraries were used as reference spectral libraries. DIA data were analyzed using DIANN (v.1.7.0). DIA-NN was set to default: Precursor FDR: 5%, Log lev: 1, Mass accuracy: 20 ppm, MS1 accuracy: 10 ppm, Scan window: 30, Implicit protein group: genes, Quantification strategy: robust LC (high accuracy). The quantification of the identified polypeptides was calculated as the average of the chromatographic fragment ion peak areas in all reference spectral libraries. Label-free, intensity-based absolute quantification (iBAQ) method was used to calculate label-free protein quantification. We calculated the peak area values of the corresponding proteins. The fraction of total (FOT) protein was used to represent the normalized abundance of a specific protein in the sample. FOT was defined as the iBAQ of a protein divided by the total iBAQ of all identified proteins in the sample. For ease of presentation, the FOT value was multiplied by 10e5, and the missing value was calculated as 10e-5.

[0063] (3) Screening for protein markers. Analyze the expression levels of protein markers and select effectively identified proteins.

[0064] In one embodiment, the number of proteins identified per single sample in 167 samples was 1,000 - 2,500. As Figure 2 shown in the cumulative curve graph, a total of 9,589 proteins were identified in 167 samples. In the present invention, the expression level of each protein in a specific sample was normalized as the proportion (fraction of total, FOT) of the expression level of all proteins in that sample. The FOT value was multiplied by 10e5 and then logarithmically processed with base 10.

[0065] (4) Establish a prediction model. The protein markers in the training set are divided into a healthy control group, a breast benign tumor group, and a breast malignant tumor, i.e., breast cancer group, and a prediction model is trained based on machine learning methods.

[0066] In one embodiment, the training set is first divided into a healthy control group, a breast benign tumor group, and a breast malignant tumor, i.e., breast cancer group, and proteins with an ANOVA test p-value less than 0.05 and an expression fold greater than two times that of the other two groups are selected. A total of 101 proteins are screened as candidate markers (ACTR3, IGHV3-53, IGHV3-33, IGHV4-34, IGHV3-23, ARPC1A, IGHV4-59, AP2A1, SSC5D, ATP2B2, IGHV4-39, PUM1, LMAN2, COPE, ARF5, SEC31A, ADD2, PARK7, IQGAP1, FNBP1L, BZW1, PPL, RPL6, ADD1, PDLIM1, CNN2, SDCBP, CTNNA1, SNX1, TRIM29, BAG3, KIF5B, PCBP1, PRDX1, ERC1, SPTAN1, EIF2A, NOP56, VASP, PTPN1, STAT1, IDH1, RDX, GAPVD1, SND1, VASN, KTN1, EEF1G, TJP1, EHD1, AFDN, HNRNPK, CTTN, CAPZA1, RPL24, TAGLN2, MAPRE1, PFN1, RAN, DBN1, PLEC, EIF4G1, APP, PON1, APOA2, APOA1, PCSK9, APOA4, APOC2, LIPC, APOC1, ANGPTL3, APOF, APOE, APOB, APOL1, APOL2, F7, LDHA, HSP90AA1, IGFBP2, ANGPTL6, MMP2, PECAM1, CYP1B1, NAA15, TYMP, PGAM4, PGK2, HLA-H, HLA-B, HLA-C, HLA-G, CTSS, ITGB1, IGLV3-25, CD44, CXADR, IGLV3-27, PVR, and HCK).

[0067] Specifically, the prediction model is established based on machine learning methods, and eight important features (STAT1, PON1, APOC1, APOC2, MMP2, IGHV4-39, IGHV3-53, and ADD2) are screened out using the XGBoost algorithm to establish the model. Table 1 shows the eight important features in the protein prediction model.

[0068] Table 1

[0069]

[0070] In one embodiment of the present invention, during the verification process of the model, the training set samples are first used for internal verification. Figure 3 is the confusion matrix for internal verification, Figure 4 is the schematic diagram of the area under the ROC curve for internal verification. As Figure 3 shown, the internal verification model includes 116 samples. The specificity of the diagnostic model is 100%, the sensitivity is 100%, the positive predictive value is 100%, and the negative predictive value is 100%. As Figure 4 shown, the ordinate represents the sensitivity value of the detection model; the abscissa represents the specificity value of the detection model, and the area under the ROC curve is 0.98.

[0071] External verification uses the validation set samples, Figure 5 is the confusion matrix of lymph node metastasis for external verification, Figure 6 is the schematic diagram of the area under the ROC curve for external verification. As Figure 5 shown, the model includes 51 samples. The specificity of the diagnostic model for diagnosing breast cancer is 83.3%, the sensitivity is 97.4%, the positive predictive value is 95%, and the negative predictive value is 90.9%. The specificity of the diagnostic model for diagnosing benign breast tumors is 100%, the sensitivity is 100%, the positive predictive value is 100%, and the negative predictive value is 100%. As Figure 6 shown, the area under the ROC curve is 0.82. This result indicates that the protein prediction model has high sensitivity and specificity and has potential application value in the clinical detection of benign and malignant breast tumors.

[0072] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. Use of a reagent for detecting 8 protein markers specific for the following breast benign tumors / breast malignant tumors in the preparation of a kit and / or chip for detecting breast benign and malignant tumors, characterized in that, the protein markers are the following 8 protein markers specific for breast benign tumors / breast malignant tumors: STAT1, PON1, APOC1, APOC2, MMP2, IGHV4-39, IGHV3-53 and ADD2; the protein markers are from exosomes, and the exosomes are serum exosomes of patients with breast benign tumors or serum exosomes of breast cancer patients; the detection is to detect the expression level of the protein markers, and the expression level is the expression level at the transcriptional level or the protein level.

2. The use according to claim 1, characterized in that, the expression level at the protein level is detected by one or more of mass spectrometry, Western blotting and ELISA methods.

3. The use according to claim 1 or 2, characterized in that, the breast malignant tumor is breast carcinoma in situ or invasive breast carcinoma.

Images