Application of protein biomarkers in the diagnosis of gastric cancer

Abstract

The present invention relates to the field of oncology. Specifically, the present invention relates to the use of protein biomarkers for the diagnosis of gastric cancer. The protein biomarkers are selected from one or a combination of two or more of GSTP1, COL15A1, VMO1, and ANGPTL2. The biomarkers of the present invention can sensitively and accurately diagnose gastric cancer.

Description

Technical Field

[0001] The present invention relates to the field of oncology, and more particularly to the application of protein biomarkers in the diagnosis and prognosis of gastric cancer. Background Art

[0002] Gastric cancer (GC) is the fifth most common malignant tumor worldwide and the fourth leading cause of cancer-related death (PMID: 33538338). Its high malignancy and poor prognosis pose a serious threat to human health. While the five-year survival rate for early-stage GC can reach 90%, the five-year survival rate for patients with advanced GC after surgery is only around 30%. Early detection, diagnosis, and treatment of GC are key to improving prognosis and reducing mortality.

[0003] Early diagnosis of gastric cancer currently relies mainly on pathological diagnosis of gastroscopy and biopsy tissue. The complicated and invasive operation causes great pain to the examinee, thus limiting its application in screening and early diagnosis in large-scale community populations.

[0004] Serological markers are minimally invasive and widely used in gastric cancer screening and early diagnosis. However, traditional serum markers such as carcinoembryonic antigen (CEA), carbohydrate antigen 72-4 (CA72-4), carbohydrate antigen 19-9 (CA19-9), and carbohydrate antigen 125 (CA125) have very limited sensitivity and specificity in the early diagnosis of gastric cancer. For example, in stage I gastric cancer, the positive rate of these traditional markers is less than 20%, and in stage III-IV gastric cancer, the positive rate is even less than 40%. Therefore, there is an urgent need for a reliable, non-invasive gastric cancer biomarker for early detection. Summary of the Invention

[0005] Based on new research findings, the present invention provides a reagent for detecting protein biomarkers for use in preparing a gastric cancer diagnosis and prognosis treatment effect evaluation kit, test paper or chip, wherein the protein biomarker is selected from one or a combination of two or more of GSTP1, COL15A1, VMO1, and ANGPTL2.

[0006] In a specific embodiment, the diagnosis refers to early diagnosis of gastric cancer patients. Compared with healthy controls, an increased level of the protein biomarker selected from the group consisting of the above-mentioned protein biomarkers indicates gastric cancer.

[0007] In a specific embodiment, the prognosis refers to the evaluation of the therapeutic effect of gastric cancer patients. Compared with before treatment, a decrease in the level of the protein biomarker selected from the above indicates the effectiveness of the treatment.

[0008] Suitable identification reagents for use in the present invention include mass spectrometry identification reagents, antibodies or antigen-binding fragments thereof, probes, or primers. In specific embodiments, the antibody is a monoclonal antibody. The present invention does not limit the species origin of the monoclonal antibody, and any antibody capable of binding to the above-mentioned protein can be used. In specific embodiments, antigen-binding fragments include, but are not limited to, Fab, Fab', (Fab')2, Fv, ScFv, bispecific antibodies, trispecific antibodies, tetraspecific antibodies, bi-scFv, and mimi antibodies. Any antibody fragment that retains antigen-binding activity is suitable for use in the present invention.

[0009] Optionally, the body fluid sample detected by the test kit, test paper or chip is selected from blood, serum, plasma, urine and saliva. Preferably, the body fluid sample is urine.

[0010] When using the kit, test paper or chip of the present invention to determine whether a subject has or is at risk of having gastric cancer, the level of a biomarker in a liquid sample from the subject is first detected, wherein the biomarker is selected from one or a combination of two or more of GSTP1, COL15A1, VMO1 and ANGPTL2; and then the level is compared with a reference value.

[0011] It should be understood that the "level" includes the absolute amount, relative amount or concentration of the biomarker and any value or parameter related thereto or that can be derived therefrom. The "comparison" generally refers to the comparison of corresponding parameters or values, such as comparing an absolute amount with an absolute reference amount, comparing a concentration with a reference concentration, or comparing the intensity signal of a biomarker obtained from a sample with an intensity signal of the same type obtained from a reference sample. The comparison can be performed manually or with the assistance of a computer. The measured or detected level of a biomarker in a sample obtained from an individual or patient and the value of the reference level can, for example, be compared to each other, and the comparison can be performed automatically by a computer program that executes an algorithm for comparison.

[0012] The "reference value" refers to a value that allows differentiation between subjects at risk of developing gastric cancer and those not at such risk, for example, a value that allows differentiation between healthy controls and gastric cancer patients. The reference value can be determined in advance and set to meet conventional requirements in terms of, for example, specificity and / or sensitivity.

[0013] In one embodiment, the reference value can be determined in one or more reference samples of a patient with gastric cancer. In another embodiment, the reference value can be determined in one or more reference samples of a patient with gastric cancer (e.g., healthy controls) from a patient who is not at risk of gastric cancer.

[0014] In certain embodiments, the reference value represents the level of the biomarker in a bodily fluid sample from a healthy individual who does not have gastric cancer. In certain embodiments, if the measured level is higher than the reference value, the subject is diagnosed with gastric cancer. A decrease in the level of the selected protein biomarker compared to pre-treatment levels indicates the effectiveness of the treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The data are the results of partial least squares discriminant analysis, where GC-pre is before gastric cancer surgery; GC-post is one week after gastric cancer surgery.

[0016] Figure 2 This is the heat map of correlation analysis of PRM quality control samples, where QC: quality control sample.

[0017] Figure 3 This is a scatter plot of PRM targeted quantitative proteomics data, where “*” represents p < 0.05; “**” represents p < 0.01; “***” represents p < 0.001; and “****” represents p < 0.0001. Figure 3 A is a scatter plot of PRM protein expression of GSTP1 in different groups; Figure 3 B is a scatter plot of PRM protein expression of GSTP1 in different groups; Figure 3 C is a scatter plot of PRM protein expression of VMO1 in different groups; Figure 3 D is a scatter plot of ANGPTL2 PRM protein expression in different groups.

[0018] Figure 4 Figure 3 is the receiver operating characteristic (ROC) curve of candidate proteins and combination markers for GC group in PRM targeted quantitative proteomics, where ROC is receiver operating characteristic (ROC) curve.

[0019] Figure 5Figure 3 ROC curves of the (GC&GC-pre) group of candidate proteins and combination markers in PRM targeted quantitative proteomics, where GC&GC-pre: gastric cancer and gastric cancer preoperative groups. DETAILED DESCRIPTION

[0020] Unless otherwise indicated, scientific and technical terms used herein have the meanings commonly understood by those skilled in the art. Furthermore, procedures in oncology, molecular genetics, nucleic acid chemistry, cell culture, biochemistry, cell biology, and the like used herein are conventional procedures widely used in the relevant fields. To facilitate a better understanding of the present invention, definitions and explanations of relevant terms are provided below.

[0021] The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention. Unless otherwise specified, the technical means used in the examples are conventional means well known to those skilled in the art.

[0022] Unless otherwise specified, the experimental methods used in the following examples are conventional methods.

[0023] Unless otherwise specified, all materials and reagents in the following examples can be obtained from commercial sources.

[0024] In the present invention, the GSTP1 is human glutathione S-transferase P1 (Glutathione S-transferase P, P09211); the COL15A1 is human type XV collagen α1 chain (Torsin-1A-interacting protein 1, P39059); the VMO1 is human vitelline membrane outer layer protein 1 homolog (Vitelline membrane outer layer protein 1 homolog, Q7Z5L0); and the ANGPTL2 is human angiopoietin-related protein 2 (Angiopoietin-related protein 2, Q9UKU9), all of which are recorded in the www.uniprot.org database.

[0025] In this study, to minimize the impact of confounding factors in the urine of different patients, a discovery group sample strategy using self-controlled samples from the same patients before and after surgery was used. Data-independent acquisition (DIA) non-targeted proteomics technology was used to compare urine proteome changes in gastric cancer patients before and one week after surgery, and to screen for significantly differentially expressed proteins between the two groups. In the validation group, the inventors used parallel reaction monitoring (PRM) targeted proteomics technology to conduct a comprehensive comparative proteomic analysis of urine from the newly included gastric cancer group, healthy controls, and the discovery group to identify potential biomarkers for the diagnosis of gastric cancer.

[0026] All urine samples in the following examples were collected from inpatients at the First Affiliated Hospital of the Air Force Medical University. Sample collection and use were approved by the hospital's ethics committee. Inclusion criteria for gastric cancer patients were a diagnosis of adenocarcinoma by endoscopic biopsy histology by at least two experienced pathologists. Patients had not undergone prior gastric cancer surgery, chemotherapy, or radiotherapy, and had no history of tumor metastasis or other malignancies. Healthy subjects were eligible if they had no tumors during their annual physical examinations and no prior history of cancer or other serious illnesses.

[0027] Example 1: Preliminary screening of urine protein markers

[0028] Liquid chromatography-high-resolution mass spectrometry (LC-MS / MS) and data independent acquisition (DIA) quantitative proteomics were used to detect the urine proteome of gastric cancer patients before and after surgery, and to screen for proteins downregulated after gastric cancer surgery.

[0029] In the mass spectrometry quantitative experiments in the following examples, HeLa standards were inserted during sample collection to perform quality control of the mass spectrometry data, and iRT-labeled peptides were added to control the retention time.

[0030] 1. Materials and Reagents

[0031] 1) Instruments: Easy-nLC 1200 (Thermo Scientific); Orbitrap QExactive HF mass spectrometer (Thermo Scientific).

[0032] 2) Main reagents: acetonitrile (Merk); precolumn (75 μm × 2 cm, C18, 3 μm, Thermo Scientific), analytical column (75 μm × 25 cm, C18, 2 μm, Thermo Scientific); iST 96X kit (PreOmics).

[0033] 3) Samples: Urine samples from 28 gastric cancer patients before and one week after surgery.

[0034] 2. Experimental Methods

[0035] 2.1 Collection of urine samples and preparation of peptide samples

[0036] (1) Fasting midstream urine was collected from gastric cancer patients before and one week after surgery. The samples were centrifuged at 1000 g for 10 min at 4°C. The supernatant was placed in a new centrifuge tube and centrifuged at 12000 g for 10 min at 4°C. After two centrifugations, the precipitate was removed and the supernatant was aliquoted into Eppendorf 1.5 mL centrifuge tubes.

[0037] (2) Take 300 μl of each of the 56 samples in step (1) and place them in Eppendorf 2.0 mL centrifuge tubes. Add 5 times the volume of -20°C pre-cooled acetone, mix well, and place at -20°C overnight for precipitation.

[0038] (3) After completing step (2), centrifuge at 12,000 g for 30 min at 4°C and discard the supernatant. Subsequent operations were performed according to the optimized steps of the iST 96X kit from Preomics.

[0039] (4) Add 50 μl of LYSE reagent to dissolve the protein precipitate in step (3), vortex mix, and heat at 95°C and 1000 rpm for 10 min;

[0040] (5) The sample heated in step (4) was cooled to room temperature, 50 μl of the prepared DIGEST was added, and enzymatic hydrolysis was carried out at 37°C and 500 rpm for 2 h;

[0041] (6) After the enzymatic hydrolysis in step (5) was completed, 100 μl of STOP solution was added and the reaction was terminated by shaking at 500 rpm for 1 min at room temperature;

[0042] (7) Using the ADAPTER configured in the kit, install the CARTRIDGE on a 1.5 centrifuge tube and mark it. Transfer the enzymatic hydrolyzate from step (6) to the connected CARTRIDGE after instant centrifugation and centrifuge at 3800 g for 1 min.

[0043] (8) Add 200 μl of WASH0, WASH1 and WASH2 in step (7), centrifuge at 3800 g for 1 min, and discard the waste liquid;

[0044] (9) Use ADAPTER to install the CARTRIDGE in a new 1.5 ml centrifuge tube, add 100 μl of ELUTE, centrifuge at 3800 g for 1 min to elute the peptide fragments, repeat the operation once and combine the two eluates;

[0045] (10) The eluate collected in step (9) was placed in a centrifugal concentrator and vacuum-dried to obtain a dried peptide sample, which was stored at -80°C.

[0046] 2.2 Establishment of DDA spectral library

[0047] 2.2.1 Preparation of DDA samples

[0048] Liquid phase separation was performed on a mixed peptide fragment from 56 samples, and the fractions were combined to construct a spectral library. The mixed peptide fragments were first pre-separated on a high-pH reversed-phase column (4.6 mm × 250 mm, Xbridge C18, 3 mm). The peptide eluent gradient was 3% to 95% buffer B (phase B: 80% ACN with 20 mM ammonium formate, adjusted to pH = 10 with ammonia), at a flow rate of 1 mL / min, and an elution time of 66 min. One fraction was collected every minute and finally combined into 18 fractions, which were drained and dissolved in 0.1% formic acid aqueous solution. The 18 fractions, the HeLa standard used for quality control, and the 56 independent samples were then subjected to iRT (Biognosys) for retention time unification.

[0049] 2.2.2 DDA mass spectrometry acquisition and analysis

[0050] (1) 18 fractionated samples were subjected to data dependent acquisition (DDA) mode mass spectrometry identification. The DDA mode mass spectrometry identification method was as follows: 5 μl of sample was separated by a Thermo EASY-nLC 1200 liquid phase system. The mobile phase consisted of 0.1% formic acid in water (A) and 80% acetonitrile solution containing 0.1% formic acid (B). The flow rate was 0.35 μL / min and the elution time was 120 min. The eluted peptides were analyzed by an Orbitrap Q Exactive HF mass spectrometer to obtain raw files. The primary mass spectrometry data acquisition range was 350-1500 m / z with a resolution of 60,000. The secondary scan resolution was 15,000 and the HCD collision energy was 27%.

[0051] (2) The raw files were then merged and analyzed using Spectronaut 15 (Biognosys) software, and the library was constructed using the default parameters of the software. The sequence database used was the UniProt Proteome human proteome database (2020 / 02); Trypsin enzymatic digestion was set; the variable window and mass spectrometry method of the data-dependent acquisition (DIA) of the sample to be identified were set based on the information of the parent ion and fragment ion and their precise retention time, and the sample was identified by mass spectrometry.

[0052] 2.3 DIA data acquisition and analysis

[0053] (1) The mass spectrometry identification method in DIA mode is as follows: under the same liquid phase system parameters as DDA, 60 variable windows are set, the scanning range of the primary mass spectrometry acquisition is 350-1250 m / z, the resolution is 120,000. The resolution of the secondary scan is 30,000, and the HCD collision energy is 27%;

[0054] (2) DIA data analysis was performed using the default parameters of Spectronaut15 software. Based on the library construction results of the graded samples, the raw files of 56 samples and quality control samples were imported into the software for protein data retrieval and analysis. Based on the iRT peptides, the software dynamically determined the ideal extraction window according to the iRT calibration and gradient stability. Spectronaut performed automatic correction and used a local normalization strategy for data normalization. The average peak area of ​​the first three peptides with an FDR less than 1% was used for protein group quantification. For the quantitative results exported by the software, proteins with more than 50% missing data in the sample were eliminated, and then the missing values ​​of the remaining proteins were filled. Finally, the filled quantitative data were log2 transformed.

[0055] (3) The statistical analysis of the quantitative results between the two groups was performed using a two-tailed paired t-test to calculate the p-value. Based on the protein quantitative level, proteins with a fold difference (FC) greater than 1.5 or FC less than 0.67 and a statistical p-value less than 0.05 were considered differentially expressed.

[0056] 3. Results

[0057] Under the condition that the FDR of parent ion and protein identification was less than 1%, and proteins with more than 50% missing values ​​in the samples were eliminated, a total of 1667 proteins were quantified; the model was constructed using partial least squares discriminant analysis (PLS-DA) ( Figure 1), the gastric cancer preoperative group (GC-pre) and the gastric cancer postoperative group (GC-post) can be clearly distinguished; p<0.05, FC>1.5 or <0.67 were identified as differentially expressed proteins, and a total of 479 differentially expressed proteins were screened, of which 182 were downregulated after surgery; all postoperatively downregulated differentially expressed proteins were reviewed through literature review and functional analysis, and finally 54 potential biomarkers were screened for subsequent PRM validation analysis.

[0058] Example 2: Analysis and Verification of Candidate Urine Protein Markers

[0059] Using Parallel Reaction Monitoring (PRM) targeted quantitative proteomics technology, 54 protein PRM targeted validation analysis was performed on urine samples from the newly included gastric cancer group (GC), healthy control group (HC) and DIA discovery group.

[0060] In the mass spectrometry quantitative experiments in the following examples, mixed samples (Quality Control, QC) were inserted in the middle of sample collection to perform quality control of mass spectrometry data, and iRT labeled peptides were added to control retention time.

[0061] 1. Materials and Reagents

[0062] 1) Instruments: Easy-nLC 1200 (Thermo Scientific); Orbitrap QExactive HF mass spectrometer (Thermo Scientific).

[0063] 2) Main reagents: acetonitrile (Merk); precolumn (75 μm × 2 cm, C18, 3 μm, Thermo Scientific), analytical column (75 μm × 25 cm, C18, 2 μm, Thermo Scientific); iST 96X kit (PreOmics).

[0064] 3) Samples: 17 gastric cancer patients and 17 sex- and age-matched healthy controls, and 56 urine samples from the discovery group.

[0065] 2. Experimental Methods

[0066] 2.1 Collection of urine samples and preparation of peptide samples

[0067] Same method as that used in the DIA discovery group.

[0068] 2.2 Collection of PRM samples

[0069] The 90 urine samples prepared in 2.1 were analyzed in the scheduled PRM mode. The same mass spectrometer, chromatographic column, and liquid phase separation conditions were used for PRM and DIA sample collection.

[0070] Mass spectrometry parameters: MS full scan parameters were 60,000 resolution (at m / z 200), scan range 350–1200 m / z, AGC setting 3E6, and maximum ion injection time 20 ms. The MS secondary scan parameters were 30,000 resolution (at m / z 200), AGC setting 2E5, maximum ion injection time 50 ms, and NCE 27%. The mass-to-charge ratio and retention time information of the target peptides (see Table 1 below) were imported into the inclusion list.

[0071] Table 1 Mass-to-charge ratio, charge, and retention time information of 111 peptides corresponding to 54 proteins

[0072]

[0073]

[0074]

[0075]

[0076]

[0077]

[0078]

[0079]

[0080]

[0081] 2.3 PRM Data Analysis

[0082] PRM data analysis was performed using the default parameters of SpectrDive 10.8. The target protein sequence list and all raw data collected by PRM were imported into SpectrDive software, which automatically calibrated the retention time and mass window based on the iRT peptides, determined the ideal extraction window, and performed automatic peak extraction and quantitative analysis.

[0083] The FDR value for peptide identification was 1%. The results of peptide identification were further manually verified to ensure the accuracy of spectral matching and extraction windows. The quantitative values ​​of peptide precursor ions were normalized using the total ion current (TIC) signal intensity values ​​of the corresponding samples to eliminate errors caused by uneven sample loading. TIC signals were extracted using Peaks studio 10.6 software. The signals of all fragment ions of each precursor were summed to calculate the endogenous peptide signals. The target peptides were quantitatively analyzed to screen for differential proteins between different groups, which were further compared with the DIA screening results.

[0084] 2.4 Statistical analysis

[0085] For PRM data, proteins with a p < 0.05 and a FC > 1.5 or < 0.67 were considered differentially expressed. Quantitative data between two groups were analyzed using paired t-tests, and Kruskal-Wallis one-way ANOVA was used for statistical comparisons of three or more groups. These analyses were performed using GraphPad Prism 8.3. Receiver operating characteristic (ROC) curves were plotted using MedCalc 20 software, and the area under the ROC curve (AUC) values ​​were generated. Statistical significance was defined as a p value < 0.05. Pearson correlation coefficient heatmaps were analyzed using the R package "pheatmap," and Venn diagrams were generated using the R package "ggVennDiagram."

[0086] 3. Results

[0087] 3.1 PRM quality control analysis

[0088] A total of 14 mixed samples were used as QC samples to monitor the stability of the system during the detection experiment; Figure 2 As shown in Figure 3, the average Pearson correlation coefficient between the quantification of each pair of QC samples was 0.95, indicating that the system maintained a stable state and the sample reproducibility was good.

[0089] 3.2 PRM Verification Results

[0090] 3.2.1 Paired t-test results

[0091] Using PRM mass spectrometry technology, 28 pairs of urine samples from gastric cancer patients before surgery (GC-pre) and one week after surgery (GC-post) in the discovery group were verified. The FC value was calculated based on the ratio of the average expression levels of the peptides in the two groups, and the P value of each peptide between the GC-pre and GC-post groups was calculated separately by paired t-test. A total of 108 peptides were quantified in the GC-pre and GC-post groups, corresponding to 53 proteins, of which 91% of the peptides corresponding to 50 proteins had a change trend consistent with the DIA results. In addition, 73% of the peptides corresponding to 44 proteins not only had a change trend consistent with the DIA but also had FC and P values ​​that met the criteria for differentially expressed proteins.

[0092] During the PRM validation phase, 17 pairs of sex- and age-matched gastric cancer (GC) and healthy control (HC) groups were included. The FC value was calculated based on the ratio of the mean expression levels of the peptides in the two groups, and the P value of each peptide between the GC and HC groups was calculated separately using a paired t-test. A total of 110 peptides were quantified in the GC and HC groups, corresponding to 54 proteins. Among them, 48% of the peptides corresponding to 34 proteins had a change trend consistent with the DIA results. In addition, 8 peptides corresponding to 7 proteins not only had a change trend consistent with the DIA but also had FC and P values ​​that met the criteria for differentially expressed proteins.

[0093] There were six peptides corresponding to five proteins whose change trends in the two experiments were consistent with those of DIA and statistically significant. Overall, the quantitative results of these six peptides were highly stable, with good reproducibility in QC samples and highly consistent quantitative results in the two experiments.

[0094] 3.2.2 Kruskal-Wallis One-Way ANOVA Test

[0095] The PRM expression levels of the five proteins with higher credibility were analyzed, and the P values ​​of the five proteins in the HC, GC, GC-pre and GC-post groups were calculated by Kruskal-Wallis one-way variance test. The FC value was calculated based on the ratio of the average expression levels of the peptide segments in each group.

[0096] Proteins GSTP1, COL15A1, VMO1, and ANGPTL2 showed high consistency in the two PRM results and DIA results. By comparing the four sets of PRM targeted quantitative proteomics data, it was found that Figure 3 As shown, GSTP1 was significantly upregulated in the GC vs HC group, GC-pre vs GC-post group, and GC vs GC-post group. There was an upregulation trend in the GC-pre vs HC group, but no significant difference was found ( Figure 3A); COL15A1 was significantly upregulated in the GC vs HC group, GC-pre vs GC-post group, and GC-pre vs HC group. There was an upregulation trend in the GC vs GC-post group, but no significant difference was found ( Figure 3 B); VMO1 and ANGPTL2 were significantly upregulated in GC vs HC group, GC-pre vs GC-post group, GC vs GC-post group, and GC-pre vs HC ( Figure 3 C and D); in addition, these four proteins were not significant in the GC vs GC-pre group and the HC vs GC-post group.

[0097] After calculation by the two tests, the proteins and corresponding peptides that were commonly upregulated in group 1 [GC vs HC] and group 2 [GC-pre vs GC-post] with P < 0.05 are shown in Table 2.

[0098] Table 2 Four proteins and corresponding peptides significantly upregulated in gastric cancer group and gastric cancer preoperative group

[0099]

[0100] 3.3 ROC Curve Analysis

[0101] To further evaluate the diagnostic efficiency of the identified differentially expressed proteins, receiver operating characteristic (ROC) curve analysis was performed on the identified candidate proteins. In this analysis, we first analyzed the five upregulated proteins in the GC vs HC groups separately. Secondly, we combined the GC and GC-pre groups into one group, collectively referred to as GC & GC-pre, and further analyzed the data from the HC group.

[0102] 3.3.1 ROC curve analysis in gastric cancer and healthy control groups

[0103] In the GC vs HC group, the ROC curve results showed that compared with the HC group, the AUC values ​​of GSTP1, COL15A1, VMO1, and ANGPTL2 proteins in the GC group were all greater than 0.70 and showed significant differences. Among them, VMO1 had the highest diagnostic efficacy, with an AUC of 0.820, a sensitivity of 88.24%, and a specificity of 76.47%. The combination of proteins GSTP1, COL15A1, VMO1, and ANGPTL2 was called "Combined", and ROC curve analysis was performed. Figure 4 and Table 3 , its AUC was 0.931, which was higher than the AUC results of individual biomarkers, with a sensitivity of 94.12% and a specificity of 76.47%.

[0104] Table 3: ROC curve analysis of candidate proteins and combined markers of GC group in PRM targeted quantitative proteomics.

[0105]

[0106] 3.3.2 ROC curve analysis in gastric cancer and preoperative gastric cancer groups and healthy controls

[0107] In the (GC & GC-pre) vs HC group results, the ROC curve results showed that compared with the HC group, the AUC values ​​of GSTP1, COL15A1, VMO1, and ANGPTL2 proteins in the TG group were greater than 0.70 and showed significant differences. Among them, COL15A1 had the highest diagnostic efficacy, with an AUC of 0.795, a sensitivity of 64.44%, and a specificity of 94.12%. The combination of GSTP1, COL15A1, VMO1, and ANGPTL2 proteins was called "Combined", and ROC curve analysis was performed. Figure 5 As shown in Table 4, its AUC was 0.850, which was higher than the AUC results of individual biomarkers, with a sensitivity of 82.22% and a specificity of 82.35%. Combining the results of the two groups, proteins GSTP1, COL15A1, VMO1, and ANGPTL2 had better AUC values, and the combination of these four proteins showed better diagnostic efficacy.

[0108] Table 4 ROC curve analysis of (GC&GC-pre) group candidate proteins and combined markers in PRM targeted quantitative proteomics

[0109]

[0110] Based on the above research, it can be seen that the four proteins can be used as urine diagnostic markers to distinguish gastric cancer patients from normal people, with high sensitivity and specificity; and the combination of the four proteins has better diagnostic efficacy, higher sensitivity and specificity.

Claims

1. Use of a reagent for detecting a protein biomarker for preparing a gastric cancer diagnostic kit, test paper or chip, wherein the protein biomarker is selected from VMO1, or a combination of GSTP1, COL15A1, VMO1 and ANGPTL2.

2. The use according to claim 1, characterized in that Elevated levels of a selected protein biomarker compared to healthy controls are indicative of gastric cancer.

3. The use according to claim 1, wherein The reagent is a mass spectrometry identification reagent, an antibody or an antigen-binding fragment thereof.

4. The use according to claim 1, wherein The body fluid sample detected by the test kit, test paper or chip is selected from urine.

5. The use according to claim 1, characterized in that The reference value of the kit, test paper or chip represents the level of the protein biomarker in the body fluid samples of healthy people.

Images