Metabolic marker combination for ovarian cancer diagnosis and application thereof

Abstract

The invention provides a metabolic marker combination for ovarian cancer diagnosis and application thereof, and belongs to the technical field of in-vitro diagnosis. The invention provides a metabolic marker combination for ovarian cancer diagnosis. The metabolic marker combination comprises phosphatidylcholine 34: 4, threonic acid, phosphatidylcholine 32: 2, pyroglutamic acid, phosphatidylcholine 36: 6 and ceramide d38: 2. According to the metabolic marker combination, plasma samples of a healthy subject and an ovarian cancer patient are analyzed, a group of biomarkers with significant difference are screened from the plasma samples, and the metabolic marker combination is a non-invasive, non-invasive and high-accuracy metabolic marker combination for distinguishing the healthy subject and the ovarian cancer patient. The method can be used for constructing a diagnostic model for distinguishing healthy subjects from ovarian cancer patients. The metabolic marker combination disclosed by the invention can be used for improving the diagnosis and evaluation capability of ovarian cancer patients, so that early discovery, early diagnosis and early intervention can be realized clinically, and the survival rate of the ovarian cancer patients is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of in vitro diagnosis, and in particular relates to a metabolic marker combination for ovarian cancer diagnosis and application thereof. Background Art

[0002] Ovarian cancer (OC) is one of the three most common malignant tumors of the female reproductive organs and is a serious threat to women's health. The ovaries are deep in the pelvic cavity. Ovarian lesions are often asymptomatic in the early stages. When symptoms appear and patients seek medical treatment, 70% of patients are already in the late stages. Therefore, early diagnosis of ovarian cancer is of great significance for improving the prognosis of patients and improving their quality of life.

[0003] The occurrence of ovarian cancer can be controlled through screening and precancerous lesion examination and treatment. Carbohydrate antigen 125 is a serum tumor marker for ovarian cancer. However, the detection sensitivity and specificity of carbohydrate antigen 125 are low, and it fails to achieve satisfactory ovarian cancer screening results. At present, the development of highly sensitive and specific markers that can be used for early diagnosis of ovarian cancer is of great value in achieving early diagnosis of ovarian cancer. Summary of the invention

[0004] The purpose of the present invention is to provide a metabolic marker combination for ovarian cancer diagnosis and application thereof. The metabolic marker combination of the present invention has the advantages of high sensitivity and strong specificity.

[0005] The invention provides a metabolic marker combination for ovarian cancer diagnosis, comprising phosphatidylcholine 34:4, threonic acid, phosphatidylcholine 32:2, pyroglutamic acid, phosphatidylcholine 36:6 ​​and ceramide d38:2.

[0006] The present invention also provides the use of a substance for detecting the metabolic marker combination described in the above scheme in the preparation of a product for ovarian cancer diagnosis or prognosis.

[0007] Preferably, the ovarian cancer includes stage I, stage II, stage III or stage IV ovarian cancer.

[0008] The present invention also provides the use of a substance for detecting the metabolic marker combination described in the above scheme in the preparation of a product for distinguishing healthy subjects from ovarian cancer patients.

[0009] Preferably, the substances used to detect the metabolic marker combination described in the above scheme include liquid chromatography-mass spectrometry detection reagents and gas chromatography-mass spectrometry detection reagents.

[0010] The present invention also provides a reagent or a kit for diagnosing ovarian cancer, wherein the reagent or the kit comprises a substance for detecting the metabolic marker combination described in the above scheme.

[0011] Preferably, the ovarian cancer includes stage I, stage II, stage III or stage IV ovarian cancer.

[0012] The present invention also provides a system for diagnosing ovarian cancer, and reagents and / or instruments for detecting the metabolic marker combination described in the above scheme.

[0013] The present invention also provides a method for constructing a diagnostic model for ovarian cancer, comprising the following steps:

[0014] Multivariate ROC curve analysis was performed on the combination of metabolic markers described in the above scheme.

[0015] Preferably, the multivariate ROC curve analysis includes: selecting 3 / 4 samples from the modeling group data as a training set, and the remaining 1 / 4 samples as a test set and using a support vector machine to randomly iterate 1000 times, and constructing a diagnostic model for ovarian cancer by calculating the average value of the final model accuracy.

[0016] The present invention provides a metabolic marker combination for ovarian cancer diagnosis, including phosphatidylcholine 34:4, threonic acid, phosphatidylcholine 32:2, pyroglutamic acid, phosphatidylcholine 36:6 ​​and ceramide d38:2. The metabolic marker combination of the present invention is obtained by analyzing plasma samples of healthy subjects and ovarian cancer patients to screen out a group of biomarkers with significant differences. The metabolic marker combination of the present invention is a non-invasive, non-invasive, and highly accurate metabolic marker combination for distinguishing healthy subjects from ovarian cancer patients, and can be used to construct a diagnostic model for distinguishing healthy subjects from ovarian cancer patients. The metabolic marker combination of the present invention can improve the diagnostic evaluation capabilities of ovarian cancer patients, so as to achieve early detection, early diagnosis, and early intervention in the clinic, thereby improving the survival rate of ovarian cancer patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0018] Figure 1 Cluster heat map based on differential metabolites of healthy group and ovarian cancer group;

[0019] Figure 2 This is a diagram of differential metabolite pathway enrichment analysis based on the healthy group and ovarian cancer group;

[0020] Figure 3Multivariate ROC curve analysis for distinguishing healthy subjects from ovarian cancer patients based on 6 metabolic markers in the modeling group;

[0021] Figure 4 Multivariate ROC curve analysis was performed to distinguish healthy subjects from ovarian cancer patients based on 6 metabolic markers in the validation group. DETAILED DESCRIPTION

[0022] The present invention provides a metabolic marker combination for ovarian cancer diagnosis, comprising phosphatidylcholine 34:4 (PC34:4), threonic acid, phosphatidylcholine 32:2 (PC 32:2), pyroglutamic acid, phosphatidylcholine 36:6 ​​(PC 36:6) and ceramide d38:2 (SM d38:2).

[0023] In a specific implementation of the present invention, the metabolic marker combination consists of phosphatidylcholine 34:4, threonic acid, phosphatidylcholine 32:2, pyroglutamic acid, phosphatidylcholine 36:6 ​​and ceramide d38:2.

[0024] The metabolite marker combination of the present invention is to analyze the plasma samples of healthy subjects and ovarian cancer patients by using metabolomics technology, and screen out a group of biomarkers with significant differences. The metabolite marker combination of the present invention is a non-invasive, non-invasive, and highly accurate metabolite marker combination for distinguishing healthy subjects from ovarian cancer patients, and can be used to construct a diagnostic model with high sensitivity and specificity for distinguishing healthy subjects from ovarian cancer patients. Early diagnosis of ovarian cancer based on the metabolite marker combination of the present invention has the advantages of being non-invasive, non-invasive, easy to obtain samples, and high in accuracy, and can improve the diagnostic evaluation capabilities of ovarian cancer patients, so as to achieve early detection, early diagnosis, and early intervention of ovarian cancer in clinical practice, thereby improving the survival rate of ovarian cancer patients.

[0025] The present invention also provides the use of a substance for detecting the metabolic marker combination described in the above scheme in the preparation of a product for ovarian cancer diagnosis or prognosis.

[0026] In a specific implementation of the present invention, the ovarian cancer includes stage I, stage II, stage III or stage IV ovarian cancer.

[0027] The present invention also provides the use of a substance for detecting the metabolic marker combination described in the above scheme in the preparation of a product for distinguishing healthy subjects from ovarian cancer patients.

[0028] In the specific implementation of the present invention, the substances used to detect the metabolic marker combination described in the above scheme include liquid chromatography-mass spectrometry detection reagents and gas chromatography-mass spectrometry detection reagents.

[0029] The present invention also provides a reagent or a kit for diagnosing ovarian cancer, wherein the reagent or the kit comprises a substance for detecting the metabolic marker combination described in the above scheme.

[0030] The present invention also provides a system for diagnosing ovarian cancer, and reagents and / or instruments for detecting the metabolic marker combination described in the above scheme.

[0031] In the specific implementation of the present invention, the reagents used to detect the metabolic marker combination described in the above scheme include liquid chromatography-mass spectrometry detection reagents and gas chromatography-mass spectrometry detection reagents.

[0032] The present invention also provides a method for constructing a diagnostic model for ovarian cancer, comprising the following steps:

[0033] Multivariate ROC curve analysis was performed on the combination of metabolic markers described in the above scheme.

[0034] In the specific implementation process of the present invention, the multivariate ROC curve analysis includes: selecting 3 / 4 samples from the modeling group data as a training set, and the remaining 1 / 4 samples as a test set and using a support vector machine to randomly iterate 1000 times, and constructing an ovarian cancer diagnosis model by statistically calculating the average value of the final model accuracy.

[0035] In the specific implementation of the present invention, in the ROC diagnostic model constructed by the modeling group, AUC = 0.982 (sensitivity: 90.5%, specificity: 96.3%, accuracy: 93.8%, precision: 95%), indicating that the diagnostic model of ovarian cancer constructed by the metabolic marker combination of the present invention has a high diagnostic efficacy.

[0036] To further illustrate the present invention, a metabolic marker combination for ovarian cancer diagnosis and its application provided by the present invention are described in detail below in conjunction with the accompanying drawings and examples, but they should not be construed as limiting the scope of protection of the present invention.

[0037] Example 1

[0038] Step 1: Test object and sample information

[0039] 1. Subjects

[0040] 1) Inclusion criteria:

[0041] Subjects must meet all of the following inclusion criteria to be eligible for this study:

[0042] (1) Women aged ≥18 years; (2) Read and fully understand, sign the informed consent form, and be able to provide blood samples for metabolomics testing; (3) Cancer group: patients diagnosed with ovarian cancer by the gold standard method of histopathological examination and with ovarian cancer stages I to IV were included.

[0043] 2) Exclusion criteria:

[0044] Subjects who meet any of the following exclusion criteria are not eligible to participate in this study:

[0045] (1) Pregnancy or lactation; (2) Emergency or in need of rescue; (3) History of blood transfusion within 7 days before sampling; (4) People who have received organ transplantation or non-autologous (allogeneic) bone marrow or stem cell transplantation; (5) People with a history of malignant tumor within 5 years or any anti-tumor treatment before sampling; (6) People with multiple primary malignant tumors at the same time.

[0046] Step 2: Plasma sample pretreatment and metabolite detection

[0047] 1. Plasma sample pretreatment

[0048] After the plasma sample was taken out of the -80°C freezer, it was thawed on ice and then vortexed for 10 seconds. Subsequently, 100 μL of plasma was taken and placed in 1000 μL of pre-cooled methyl tert-butyl ether and methanol mixture (the volume ratio of methyl tert-butyl ether and methanol was 3:1), and vortexed to obtain the sample extract. Next, 500 μL of methanol and water mixture (the volume ratio of methanol and water was 3:1) was added to the sample extract, and centrifuged at 12700 rpm for 5 minutes to separate the upper organic phase and the lower aqueous phase.

[0049] -Aqueous phase: Take out 400 μL of hydrophilic metabolites from the lower layer and transfer them to a centrifuge tube, add 1100 μL of ice methanol, and vortex to precipitate the protein; after protein precipitation, incubate at 4°C for 1 hour, then centrifuge at 12700 rpm for 10 minutes at 4°C, take out 1000 μL of the supernatant and transfer it to the corresponding polar centrifuge tube, then take out 350 μL of the supernatant and transfer it to the corresponding numbered GC tube, and then spin dry it in a vacuum concentrator (Speed-Vac).

[0050] 1) Add 200 μL of mass spectrometry grade water to the dried polar centrifuge tube for reconstitution and incubate at room temperature for 15 min; after incubation, vortex the centrifuge tube, ultrasonically treat for 5 min, and then centrifuge the centrifuge tube for 5 min (12000 rpm) at room temperature; take 180 μL of supernatant from the centrifuge tube into a 2 mL glass injection bottle for polar metabolite detection (LC-MS).

[0051] 2) The dried GC tube sample was dissolved in 10 μL of methoxyamine hydrochloride (MeOX) solution, vortexed for 10 seconds, and the centrifuge tube was centrifuged at 12700 rpm for 1 minute at room temperature; it was transferred to a heating mixer, set at 30°C, 500 rpm, and incubated for 90 minutes; 90 μL of N-methyl-N-(trimethylsilyl)trifluoroacetamide (MSTFA) solution was added, and it was transferred to a heating mixer, set at 37°C, 500 rpm and incubated for 30 minutes; the centrifuge tube was centrifuged at 12700 rpm for 1 minute at room temperature, and 85 μL of the supernatant was transferred to the corresponding injection bottle, and detected by the machine (GC-MS);

[0052] -Organic phase: 350 μL of organic phase was taken from the upper layer and transferred to a centrifuge tube, and then dried using a vacuum concentrator (Speed-Vac). The dried sample was reconstituted with 200 μL of acetonitrile / isopropanol (v / v, 3:1) and incubated at room temperature for 15 min; after incubation, the centrifuge tube was vortexed and treated with ultrasound (ultrasonic treatment at room temperature) for 5 min, and then the centrifuge tube was centrifuged at room temperature for 5 min (12000 rpm); 180 μL of supernatant was taken from the centrifuge tube and transferred to a 2 mL glass injection vial, which was a non-polar substance and was detected by LC-MS.

[0053] 2. High-throughput gas chromatography-mass spectrometry (GC-MS) detection:

[0054] -GC-MS parameters: After derivatization, the sample is injected into a -5Sil MS column (Restek 30m*0.25mm, 0.25μm) gas chromatography system (7890B, Agilent) and time-of-flight mass spectrometry system (PEGASUS-BT, Leco) were used for detection. High-purity helium was used as carrier gas with a flow rate of 1mL / min. The gas chromatography heating program was: initial temperature 50℃, maintained for 1min, and continuously heated to 330℃ at 15℃ / min, maintained for 2min. The main parameters of mass spectrometry were: injection port temperature 280℃, electron ionization source (EI) temperature 300℃, electron energy 70eV; scanning range 50~500amu, detection voltage 1.2kV.

[0055] 3.UHPLC-MS detection:

[0056] (1) Polar metabolite chromatographic parameters: Polar metabolite analysis was performed using Waters ACQUTTY HSS T3 chromatographic column (1.8 μm 2.1*100 mm, 1.8 μm) was used for small molecule separation, and the column temperature was 40°C; the liquid chromatography and mass spectrometer used ACQUITYUPLC I-Class liquid chromatography system (Waters) and Q-Exactive mass spectrometry system (Thermo Fisher Scientific), respectively; the mobile phase composition was as follows: mobile phase A was an aqueous solution containing 0.1% formic acid; mobile phase B was an acetonitrile solution containing 0.1% formic acid, and the flow rate was 0.4 mL / min; the separation elution gradient was as follows: 0-1 min, 1% mobile phase B; 11-13 min, 40%-70% mobile phase B, 13-15 min, 70%-99% mobile phase B; 15-18 min, 99% mobile phase B; 19 min, 1% mobile phase B; the sample injection volume was 3 μL, and the autosampler temperature was 10°C;

[0057] (2) Non-polar metabolite chromatography parameters: Lipid metabolite analysis using Waters ACQUTTY A BEH C8 (2.1*100 mm, 1.7 μm) chromatographic column was used for small molecule separation, and the column temperature was 60°C. The liquid chromatography and mass spectrometry used ACQUITY UPLC I-Class liquid chromatography system (Waters) and Q-Exactive mass spectrometry system (Thermo Fisher Scientific), respectively. The mobile phase composition was as follows: mobile phase A was water; mobile phase B was acetonitrile / isopropanol (7 / 3, V / V), both containing 0.1% acetic acid and 1% ammonium acetate, and the flow rate was 0.4 mL / min. The separation elution gradient was as follows: 0-1 min, 55% mobile phase B; 1-4 min, 55%-75% mobile phase B; 4-12 min, 75%-89% mobile phase B; 12-15 min, 89%-100% mobile phase B; 15-19.5 min, 100% mobile phase B; 19.51-24 min, 55% mobile phase B; the injection volume was 2 μL, and the autosampler temperature was 10°C;

[0058] (3) LC-MS mass spectrometry parameters: Full scan and data-dependent acquisition (DDA) were used to obtain primary (MS1) and secondary mass spectrometry (MS2) spectra for both polar and non-polar metabolites. The full scan mass spectrum range was 100-1500 Da. The secondary scan mode (Full MS / dd-MS2) scan range was 100-310, 300-710 and 700-1500 Da. The MS instrument used an Orbitrap high-resolution mass spectrometer (Thermo Fisher) equipped with an electrospray ionization source (ESI), and data were collected in positive and negative ionization modes. The specific parameters are as follows: automatic gain control (AGC) is 3E+6, Maximum IT is 200ms, the first-level full scan (Full scan) resolution is 70,000FWHM (@200m / z), in the second-level scan mode (Full MS / dd-MS2), the resolution of the secondary mass spectrometer is 17,500, the quadrupole window is 1.5m / z, the automatic gain control (AGC) is 1E+5, the maximum ion injection time is 50ms, and the relative collision energy (HCD) is 30eV; the ion source voltage is +3,500V in positive ion mode and -3,500V in negative ion mode, the nebulizer is 20psi, the sheath gas temperature is 400℃, and the sheath gas flow rate is 10L / min.

[0059] Step 3: Metabolomics data preprocessing and metabolite identification

[0060] 1. Metabolomics data preprocessing:

[0061] The metabolomics data preprocessing steps include peak extraction, peak alignment, peak filtering, and missing value filling. First, the original mass spectrometry data is subjected to peak extraction, peak alignment, and peak filtering to reduce the systematic data interference caused by instrument detection time fluctuations. The polar metabolite detection platform and the lipid metabolite detection platform are processed separately to avoid mutual interference between different detection platforms. The retention time of the data set is calibrated according to the historical detection of the instrument. After calibration, different filtering criteria are applied to remove the following interfering peaks: (i) isotope peaks; (ii) fragments caused by the ionization method that cause the analyte to be broken into fragments in the source; (iii) redundant peaks, such as additional low-intensity adducts of the same analyte and redundant derivatives, to ensure the quality of the analyzable data set, and to remove the missing values ​​of characteristic peaks > 50% in all samples. The random forest chain equation (MICEForest) is used to fill the missing values. The processed data matrix is ​​used as the raw data for statistical analysis, and a series of data preprocessing such as homogenization and standardization can be further performed to obtain the final mass spectrometry matrix data.

[0062] 2. Identification of metabolites:

[0063] The original data is analyzed by the software to obtain the spectral information of the compound parent ion and secondary fragment ion, such as the mass-to-charge ratio (m / z) of the primary mass spectrum, the fragments of the secondary ion, and the retention time. The metabolites are qualitatively identified by matching the spectral information of the primary and secondary metabolites in the database. Commonly used metabolite databases include HMDB ( www.hmdb.ca )、PubChem( https: / / pubchem.ncbi.nlm.nih.gov )、MassBank( http: / / www.massbank.jp )、MassBank of North America( https: / / massbank.us ), and the lipid database Lipidmap ( https: / / www.lipidmaps.org ) and Lipidblast( https: / / fiehnlab.ucdavis.edu / projects / LipidBl ast ); Based on the metabolites identified in the relevant database, the metabolites are finally verified according to the retention time, MS1, and MS2 mass spectrometry information when the standard is separated under the same chromatographic column and mass spectrometry conditions. The standard for metabolite identification is that the retention time is within 0.1min deviation and the mass accuracy deviation of the metabolite is less than 10ppm.

[0064] Step 4: Diagnostic method for distinguishing healthy group and ovarian cancer group based on metabolomic characteristics

[0065] 1. Subject information

[0066] A total of 218 subjects were recruited in this study, from plasma samples collected from two medical centers. Including healthy subjects (healthy group) and ovarian cancer patients (ovarian cancer group). The specific distribution is as follows: There are 145 plasma samples in the modeling group, including 82 healthy subjects and 63 ovarian cancer patients; there are 73 plasma samples in the validation group, including 41 healthy subjects and 32 ovarian cancer patients (Table 1). Plasma samples were collected in the early morning on an empty stomach, and all collected plasma samples were stored in a -80°C refrigerator.

[0067] Table 1. Sample information

[0068] Modeling Group Validation Group Number of people Total number (N=218) 145 73 Grouping Healthy subjects 82 41 Ovarian cancer patients 63 32 age Mean(Standard Deviation) 49.8±12.8 51.6±11.9

[0069] 2. Screening of metabolic marker combinations

[0070] (1) Data standardization processing:

[0071] The metabolomics data were processed by Log-transformation and NormAE in turn. Log-transformation aims to scale the original data to a specific interval and eliminate the order of magnitude differences between different samples. NormAE aims to eliminate the batch effects between different samples and obtain standardized data for subsequent multivariate statistical analysis.

[0072] (2) Correlation analysis of differential metabolites:

[0073] First, in the healthy subjects and ovarian cancer patient samples of the modeling group, the orthogonal partial least squares discriminant analysis (OPLS-DA) method was used to distinguish the healthy group and ovarian cancer group samples. The variable importance projection (VIP value) was used to measure the influence and explanatory power of the content of each metabolite component on the classification and discrimination of samples, and to assist in the screening of metabolic markers. The VIP value was greater than 1 as a condition, and the metabolites that met the false discovery rate (FDR) between the two groups was less than 0.05 were selected as differential metabolites. A total of 177 differential metabolites between the healthy group and the ovarian cancer group were screened out, of which 114 metabolites were upregulated in the healthy group, such as linoleic acid, fumaric acid, arginine, histidine and tryptophan; 63 metabolites were upregulated in the ovarian group, such as kynurenine, glutamic acid, leucine, palmitoleic acid and valine ( Figure 1 ).

[0074] Secondly, these 177 differential metabolites were further subjected to KEGG pathway enrichment analysis. Figure 2 As shown in the data, there were significant differences (P-value < 0.05) between the healthy group and the ovarian cancer group in seven pathways, including valine, leucine and isoleucine biosynthesis, arginine biosynthesis, alanine, aspartate and glutamate metabolism, arginine and proline metabolism, glyoxylate and dicarboxylic acid metabolism, valine, leucine and isoleucine degradation, and glutathione metabolism. These differences may reflect the specific utilization of these metabolites by tumor cells, and also suggest that these pathways play an important role in the occurrence and development of ovarian cancer.

[0075] (3) Screening of metabolic marker combinations:

[0076] In order to further find the metabolic marker combinations with significant differences in the plasma samples of the healthy group and the ovarian cancer group, multivariate statistical analysis methods were used for screening. Specifically, the univariate ROC curve AUC value and multiple hypothesis test FDR (error detection rate) and the VIP value in the orthogonal partial least squares discriminant analysis (OPLS-DA) model were combined to screen out potential biomarkers.

[0077] 1) First, univariate ROC analysis is a method to examine the impact of a single metabolite on the performance of a classification model, and to explore the evaluation of a single feature on the diagnostic performance of the ROC model. Based on the samples of healthy subjects and ovarian cancer patients in the modeling group, univariate ROC curve analysis was performed on the samples of the healthy group and the ovarian cancer group to obtain the area under the ROC curve (AUC) values ​​of all single metabolites. The univariate ROC curve AUC value > 0.8 was used as the screening condition, and the metabolites that simultaneously met the FDR < 0.05 between the two groups;

[0078] 2) Secondly, based on the multivariate statistical analysis in the OPLS-DA model, the VIP values ​​of all variables contributing to the model are obtained. The VIP value is used to measure the influence and explanatory power of the content of each metabolite component on the classification and discrimination of samples, and to assist in the screening of metabolic markers. Based on the metabolomics data, an OPLS-DA model is constructed between the healthy subjects and ovarian cancer patient samples in the modeling group, and between the healthy group and the ovarian cancer group, to obtain the VIP value of each metabolite contributing to the model, and VIP>2 is used as the screening condition;

[0079] 3) Based on the univariate AUC>0.8, FDR<0.05 and VIP>2 as the screening conditions, the intersection was taken. Finally, a total of 6 metabolome features were screened out. The combination of 6 metabolite markers was phosphatidylcholine 34:4 (PC 34:4), threonic acid, phosphatidylcholine 32:2 (PC 32:2), pyroglutamic acid, phosphatidylcholine 36:6 ​​(PC 36:6), ceramide d38:2 (d20:1 / 18:1) (SM d38:2 (d20:1 / 18:1)) (Table 2). Subsequently, the diagnostic model of healthy subjects and ovarian cancer patients was constructed using machine learning methods.

[0080] Table 2. Combination of 6 metabolic markers to distinguish healthy group from ovarian cancer group

[0081] Serial number Metabolic Markers-English Metabolic Markers-Chinese AUC VIP 1 PC34:4 Phosphatidylcholine 34:4 0.895 2.323 2 Threonicacid Threonic acid 0.882 2.165 3 PC32:2 Phosphatidylcholine 32:2 0.853 2.146 4 Pyroglutamic acid Pyroglutamate 0.851 2.088 5 PC36:6 Phosphatidylcholine 36:6 0.842 2.026 6 SMd38:2(d20:1 / 18:1) Ceramide d38:2 (d20:1 / 18:1) 0.835 2.007

[0082] 3. Build a diagnostic model to distinguish healthy subjects from ovarian cancer patients

[0083] In order to evaluate the diagnostic effect of the six selected metabolite marker combinations in distinguishing the healthy group from the ovarian cancer group, a multivariate ROC curve analysis was performed for the six metabolite marker combinations. Specifically, 3 / 4 samples were randomly selected from the modeling group data as the training set (training), and the remaining 1 / 4 samples were used as the test set (test); the training set was used to build and train the machine learning classification model, and the remaining samples were used as the test set to verify the discriminative ability of the trained model. The machine learning algorithm (support vector machine, SVM) was randomly iterated 1000 times, and a diagnostic model for distinguishing healthy subjects from ovarian cancer patients was constructed by statistically calculating the average value of the final model accuracy.

[0084] Generally speaking, the ROC curve is a method to study the relationship between model sensitivity and specificity, with sensitivity as the ordinate and 1-specificity as the abscissa. The evaluation basis is to compare the AUC value of the area under the curve. When AUC is greater than 0.5, the closer it is to 1, the better the model performance and the better the judgment effect; if it is less than 0.5, it means that the performance of the model is close to random guessing, with poor discrimination ability and poor accuracy. In addition to common parameters such as the receiver operating curve (ROC) and the area under the curve (AUC), the ROC classification prediction model also includes sensitivity, specificity, accuracy, and precision.

[0085] The sensitivity is:

[0086]

[0087] Specificity is:

[0088]

[0089] The accuracy is:

[0090]

[0091] The accuracy is:

[0092]

[0093] in:

[0094] TP (True positive): The number of samples that are actually positive examples that are correctly predicted as positive examples;

[0095] TN (Ture Negative): The number of samples that are actually negative examples but are correctly predicted as negative examples;

[0096] FP (False Positive): The number of samples that are actually negative examples but are mistakenly predicted as positive examples;

[0097] FN (False Negative): The number of samples that are actually positive examples but are mistakenly predicted as negative examples;

[0098] The results are as follows Figure 3 As shown, in the ROC diagnostic model constructed by the modeling group, AUC = 0.982 (sensitivity: 90.5%, specificity: 96.3%, accuracy: 93.8%, precision: 95%), indicating that the diagnostic model for healthy subjects and ovarian cancer patients constructed based on the combination of 6 metabolic markers has a high diagnostic efficacy.

[0099] 4. External Validation of the Diagnostic Model in Healthy Subjects and Ovarian Cancer Patients

[0100] In order to further verify the predictive effect of the diagnostic model for healthy subjects and ovarian cancer patients constructed based on the combination of 6 metabolite markers. The validation group data set is used as an unknown sample and put into the diagnostic model of healthy subjects and ovarian cancer patients constructed using the modeling group to evaluate the independent validation effect of the model on the unknown data set other than the modeling group data set. Based on the metabolomics data of 6 metabolites, each sample will output a corresponding probability value (Probability). According to the probability value of each sample as the diagnostic threshold, a set of confusion matrices (including true positive, true negative, false positive, and false negative) are obtained, and the sensitivity and specificity can be calculated according to the formula. In the ROC analysis diagram with sensitivity as the ordinate and 1-specificity as the abscissa, the values ​​corresponding to different sensitivities and 1-specificities have their corresponding points. Similarly, when the probability value of each sample is used as the diagnostic threshold, multiple different points can be obtained from the ROC analysis diagram. Linking these points will draw a ROC curve diagram.

[0101] Among them, the point with the best sensitivity and specificity was selected, and the diagnostic threshold at this time was 0.4489. As shown in the confusion matrix results in the following table (Table 3), in the diagnostic model constructed based on the above 6 metabolic markers, with 0.4489 as the diagnostic threshold, among the 32 healthy subjects, 28 were judged as the healthy group, and 4 were misjudged as the ovarian cancer group; among the 41 ovarian cancer patients, 39 were diagnosed as the ovarian cancer group, and 2 were misjudged as the healthy group. The ROC analysis results of the diagnostic model in the validation group are shown in the figure below. Figure 4As shown, the sensitivity and specificity were calculated based on the results of the confusion matrix, and the results showed that AUC = 0.978 (sensitivity: 87.5%, specificity: 95.1%, accuracy: 91.8%, precision: 93.3%), indicating that the diagnostic model for healthy subjects and ovarian cancer patients established by the combination of the screened 6 metabolic markers has good predictive ability.

[0102] Table 3. Confusion matrix of the diagnostic model for distinguishing healthy subjects from ovarian cancer patients in the validation group

[0103] Health Group Ovarian cancer group Healthy group, N=32 28 (true positive) 4 (false negative) Ovarian cancer group, N=41 2 (false positive) 39 (true negative)

[0104] Although the above embodiment describes the present invention in detail, it is only a part of the embodiments of the present invention, not all of the embodiments. People can also obtain other embodiments based on this embodiment without creativity, and these embodiments all fall within the protection scope of the present invention.

Claims

1. A metabolic marker combination for ovarian cancer diagnosis, characterized in that: Including phosphatidylcholine 34:4, threonic acid, phosphatidylcholine 32:2, pyroglutamic acid, phosphatidylcholine 36:6 ​​and ceramide d38:

2.

2. Use of a substance for detecting the metabolic marker combination according to claim 1 in the preparation of a product for ovarian cancer diagnosis or prognosis.

3. The use according to claim 2, characterized in that: The ovarian cancer includes stage I, stage II, stage III or stage IV ovarian cancer.

4. Use of a substance for detecting the combination of metabolic markers according to claim 1 in the preparation of a product for distinguishing healthy subjects from ovarian cancer patients.

5. The use according to any one of claims 2 to 4, characterized in that: The substance used to detect the metabolic marker combination according to claim 1 includes a liquid chromatography-mass spectrometry detection reagent and a gas chromatography-mass spectrometry detection reagent.

6. A reagent or kit for diagnosing ovarian cancer, characterized in that: The reagent or kit comprises substances for detecting the metabolic marker combination according to claim 1.

7. The reagent or kit according to claim 6, characterized in that The ovarian cancer includes stage I, stage II, stage III or stage IV ovarian cancer.

8. A system for ovarian cancer diagnosis, characterized in that A reagent and / or instrument for detecting the metabolic marker combination according to claim 1.

9. A method for constructing a diagnostic model for ovarian cancer, characterized in that: The following steps are involved: A multivariate ROC curve analysis was performed on the combination of metabolite markers described in claim 1.

10. The method according to claim 9, characterized in that The multivariate ROC curve analysis includes: selecting 3 / 4 samples from the modeling group data as a training set, and the remaining 1 / 4 samples as a test set; using a support vector machine to randomly iterate 1000 times; and constructing an ovarian cancer diagnosis model by calculating the average value of the final model accuracy.

Images