A plasma metabolic marker composition, screening method and application thereof

Abstract

The application discloses a plasma metabolic marker composition and a screening method and application thereof, and relates to the technical field of biomedicine. The plasma metabolic marker composition comprises glycyl-L-glutamic acid, phosphatidylcholine 38:7, pyrrole-2-formic acid and phosphatidylethanolamine 36:2p. The metabolic marker composition is significantly different in different cancer types, can distinguish different cancer types of cancer patients, and realizes the detection of pan-cancer. The metabolic marker composition is non-invasive and minimally invasive when used for the detection of pan-cancer, is more suitable for the screening and detection of a large number of ordinary people, has high sensitivity and specificity, is more suitable for the early screening of cancer, and is helpful for the early screening, early diagnosis and early treatment of cancer, and is helpful for improving the survival rate and cure rate of cancer patients.

Description

Technical Field

[0001] This invention relates to the field of biomedical technology, and in particular to a plasma metabolic biomarker composition, its screening method, and its application. Background Technology

[0002] Cancer refers to the multi-stage process by which normal cells transform into tumor cells, typically progressing from precancerous lesions to malignant tumors. Common types of cancer include lung cancer, colorectal cancer, stomach cancer, breast cancer, liver cancer, and prostate cancer, which have become major threats to human health. Currently, common clinical methods for cancer screening and diagnosis include pathological examination, imaging examination, and molecular screening.

[0003] Pathological examination, as the gold standard for cancer diagnosis, is now widely used in clinical practice and scientific research. Firstly, it clarifies and verifies the preoperative diagnosis, improving the level of clinical diagnosis. Secondly, once the diagnosis is confirmed, it can determine further treatment plans and estimate prognosis, thereby improving the level of clinical treatment. Furthermore, it yields a wealth of valuable research data. However, because pathological examination is usually invasive, patient compliance is low, making it unsuitable for large-scale early screening.

[0004] Imaging examinations can provide images of the body's internal structure, revealing the location and size of lesions, which helps doctors diagnose diseases, determine the severity of conditions, and provide post-diagnosis monitoring for patients. However, due to their radiation exposure and the fact that lesions are only detectable when they reach a certain size, they are not suitable for early cancer screening.

[0005] Currently, commonly used non-invasive tumor markers in molecular screening include CEA and CA19-9. However, their sensitivity or specificity in cancer diagnosis is not high, and the test results are easily affected by various factors such as diet, medication, and lifestyle habits. Therefore, there is an urgent need to find novel tumor markers to promote early intervention and treatment of cancer and prolong patient survival.

[0006] Metabolic biomarkers, as a novel type of tumor marker that has gained considerable attention in recent years, are primarily based on metabolomics methods. They reveal the overall metabolic changes under the influence of intrinsic and extrinsic factors to reflect a series of biological events occurring in a pathophysiological process. Identifying disease metabolic biomarkers has been one of the most widely applied areas of metabolomics. Although this approach requires highly specialized technology and equipment, involves complex data analysis, and has a relatively low degree of standardization, it can improve diagnostic accuracy and has the potential to become a tool for early cancer diagnosis and prognostic assessment.

[0007] Therefore, existing technologies still need to be improved and developed. Summary of the Invention

[0008] In view of the shortcomings of the prior art, the purpose of this invention is to provide a plasma metabolic biomarker composition and its screening method and application, which aims to solve the problems of existing cancer diagnosis methods being invasive or having low accuracy.

[0009] The technical solution of the present invention is as follows:

[0010] In a first aspect, the present invention provides a plasma metabolic marker composition comprising glycyl-L-glutamic acid, phosphatidylcholine 38:7, pyrrole-2-carboxylic acid and phosphatidylethanolamine 36:2p.

[0011] Optionally, the plasma metabolic marker composition further includes at least one of hypoxanthine, phosphatidylethanolamine 34:2p, phosphatidylethanolamine 36:1p, and pyroglutamic acid.

[0012] Optionally, the plasma metabolic marker composition further includes at least one of allantoic acid, phosphatidylethanolamine 36:3p, arabinose, oxalic acid, and phosphatidylcholine 34:2e.

[0013] Optionally, the plasma metabolic marker composition further includes at least one of L-malic acid, nicotinic acid, oxalic acid, phosphatidylethanolamine 38:5p, and uridine.

[0014] Optionally, the plasma metabolic marker composition comprises glycyl-L-glutamic acid, phosphatidylcholine 38:7, pyrrole-2-carboxylic acid, phosphatidylethanolamine 36:2p, hypoxanthine, phosphatidylethanolamine 34:2p, phosphatidylethanolamine 36:1p, pyroglutamic acid, allantoic acid, phosphatidylethanolamine 36:3p, arabinose, oxalic acid, and phosphatidylcholine 34:2e.

[0015] A second aspect of the present invention provides a method for screening the plasma metabolic marker composition of the present invention as described above, comprising the following steps:

[0016] Plasma from patients with gastric cancer, colorectal cancer, and lung cancer was extracted to obtain extracts;

[0017] The organic and aqueous phases in the extract were detected using chromatography-mass spectrometry to obtain detection data;

[0018] The detection data is processed and then the metabolites are identified to obtain metabolomics data.

[0019] The metabolomics data were analyzed to screen and obtain the plasma metabolic biomarker composition.

[0020] Optionally, the step of analyzing the metabolomics data and screening to obtain the plasma metabolic biomarker composition specifically includes:

[0021] The metabolomics data were subjected to linear correlation analysis, and the Pearson correlation coefficient was calculated. The plasma metabolic biomarker composition was obtained by screening with the absolute value of the Pearson correlation coefficient being greater than a preset value.

[0022] Optionally, the absolute value of the Pearson correlation coefficient is greater than or equal to 0.3 as a screening criterion.

[0023] A third aspect of the invention provides the use of the plasma metabolic marker composition of the invention as described above in the preparation of products for detecting pan-cancer species.

[0024] Optionally, the pan-cancer species include gastric cancer, colorectal cancer, and lung cancer.

[0025] Beneficial Effects: The metabolic biomarker composition described in this invention shows significant differences in different cancer types (such as gastric cancer, colorectal cancer, and lung cancer), enabling the differentiation of different cancer types in cancer patients and achieving pan-cancer detection. Pan-cancer detection using the metabolic biomarker composition is non-invasive and minimally invasive, making it more suitable for large-scale screening and detection in the general population. Furthermore, this metabolic biomarker composition has high sensitivity and specificity, making it more suitable for early cancer screening, which helps in early screening, early diagnosis, and early treatment of cancer, and helps improve the survival and cure rates of cancer patients. Attached Figure Description

[0026] Figure 1 This is a linear correlation curve of hypoxanthine in the three cancer modeling groups in Example 1.

[0027] Figure 2 This is a linear correlation curve of glycyl-L-glutamic acid in three cancer modeling groups in Example 1.

[0028] Figure 3 This is a linear correlation curve of allantoin in the three cancer modeling groups in Example 1.

[0029] Figure 4 This is a linear correlation curve of phosphatidylethanolamine 36:3p in three cancer modeling groups in Example 1.

[0030] Figure 5 This is a linear correlation curve of arabinose in the three cancer modeling groups in Example 1.

[0031] Figure 6 The graph shows the linear correlation curves of phosphatidylethanolamine 34:2p in three cancer modeling groups in Example 1.

[0032] Figure 7 This is a linear correlation curve of L-malic acid in three cancer modeling groups in Example 1.

[0033] Figure 8 This is a linear correlation curve of oxalic acid in the three cancer modeling groups in Example 1.

[0034] Figure 9 This is a linear correlation curve of phosphatidylethanolamine 36:1p in three cancer modeling groups in Example 1.

[0035] Figure 10 This is a linear correlation curve of niacin in the three cancer modeling groups in Example 1.

[0036] Figure 11 The graph shows the linear correlation curves of phosphatidylcholine 38:7 in the three cancer modeling groups in Example 1.

[0037] Figure 12 This is a linear correlation curve of pyroglutamic acid in the three cancer modeling groups in Example 1.

[0038] Figure 13 This is a linear correlation curve of oxalic acid in the three cancer modeling groups in Example 1.

[0039] Figure 14 This is a linear correlation curve of phosphatidylethanolamine 38:5p in three cancer modeling groups in Example 1.

[0040] Figure 15 This is a linear correlation curve of uridine in the three cancer modeling groups in Example 1.

[0041] Figure 16 The graph shows the linear correlation curves of phosphatidylcholine 34:2e in the three cancer modeling groups in Example 1.

[0042] Figure 17 The graph shows the linear correlation curves of pyrrole-2-carboxylic acid in the three cancer modeling groups in Example 1.

[0043] Figure 18 The graph shows the linear correlation curves of phosphatidylethanolamine 36:2p in three cancer modeling groups in Example 1.

[0044] Figure 19 This is a multivariate multiclass ROC curve analysis of 18 metabolic markers in the plasma samples of the modeling group in Example 2.

[0045] Figure 20This is a multivariate multiclass ROC curve analysis of 18 metabolic markers in the plasma samples of the validation group in Example 2. Detailed Implementation

[0046] This invention provides a plasma metabolic biomarker composition, its screening method, and its application. To make the objectives, technical solutions, and effects of this invention clearer and more explicit, the invention is further described in detail below. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0047] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0048] The following explanation addresses some of the metabolic markers that appear in the embodiments described below.

[0049] Phosphatidylcholine belongs to the glycerophospholipid class. One of its three side chains is phosphatidylcholine, and the other two are fatty acids.

[0050] In phosphatidylcholine 38:7, 38:7 refers to the fact that the side chains of the other two fatty acids contain 38 carbon atoms and 7 double bonds.

[0051] In phosphatidylcholine 34:2e, 34:2 refers to the fact that the other two fatty acid side chains contain 34 carbon atoms and 2 double bonds; e represents one of the two fatty acid side chains, where the covalent ester bond originally formed by the esterification of one hydroxyl group and one fatty acid molecule has been replaced by an ether bond, thus resulting in one less oxygen atom.

[0052] Phosphatidylethanolamine belongs to the glycerophospholipid class. One of its three side chains is phosphatidylethanolamine, while the other two are fatty acids.

[0053] In phosphatidylethanolamine 34:2p, 34:2 refers to the fact that the other two fatty acid side chains contain 34 carbon atoms and 2 double bonds; p represents one of the two fatty acid side chains. The covalent ester bond originally formed by the esterification of one hydroxyl group and one fatty acid molecule has been replaced by an ether bond, thus resulting in one less oxygen atom and one more double bond.

[0054] In phosphatidylethanolamine 36:1p, 36:1 refers to the fact that the other two fatty acid side chains contain 36 carbon atoms and 1 double bond; p represents one of the two fatty acid side chains. The covalent ester bond originally formed by the esterification of 1 hydroxyl group and 1 fatty acid molecule has been replaced by an ether bond, thus resulting in one less oxygen atom and one more double bond.

[0055] In phosphatidylethanolamine 36:2p, 36:2 refers to the fact that the other two fatty acid side chains contain 36 carbon atoms and 2 double bonds; p represents one of the two fatty acid side chains. The covalent ester bond originally formed by the esterification of one hydroxyl group and one fatty acid molecule has been replaced by an ether bond, thus resulting in one less oxygen atom and one more double bond.

[0056] In phosphatidylethanolamine 36:3p, 36:3 refers to the fact that the other two fatty acid side chains contain 36 carbon atoms and 3 double bonds; p represents one of the two fatty acid side chains. The covalent ester bond that was originally formed by the esterification of one hydroxyl group and one fatty acid molecule has been replaced by an ether bond, so there is one less oxygen atom and one more double bond.

[0057] In phosphatidylethanolamine 38:5p, 38:5 refers to the fact that the other two fatty acid side chains contain 38 carbon atoms and 5 double bonds; p represents one of the two side chains. The covalent ester bond, which was originally formed by the esterification of one hydroxyl group and one fatty acid molecule, has been replaced by an ether bond, thus resulting in one less oxygen atom and one more double bond.

[0058] This invention provides a plasma metabolic marker composition comprising glycyl-L-glutamic acid, phosphatidylcholine 38:7, pyrrole-2-carboxylic acid, and phosphatidylethanolamine 36:2p.

[0059] In this embodiment, the plasma metabolic biomarker composition containing four plasma metabolic biomarkers shows significant differences across different cancer types, enabling the differentiation of cancer patients into different cancer types (such as gastric cancer, colorectal cancer, and lung cancer), achieving pan-cancer detection (including gastric cancer, colorectal cancer, and lung cancer). Using this metabolic biomarker composition for pan-cancer detection is non-invasive and minimally invasive, making it more suitable for large-scale screening and detection in the general population. Furthermore, this metabolic biomarker composition has high sensitivity and specificity, making it more suitable for early cancer screening, contributing to early cancer screening, early diagnosis, and early treatment, and helping to improve the survival and cure rates of cancer patients.

[0060] In some embodiments, the plasma metabolic marker composition further includes at least one of hypoxanthine, phosphatidylethanolamine 34:2p, phosphatidylethanolamine 36:1p, and pyroglutamic acid. That is, in this embodiment, the plasma metabolic marker composition includes glycyl-L-glutamic acid, phosphatidylcholine 38:7, pyrrole-2-carboxylic acid, phosphatidylethanolamine 36:2p, and at least one of hypoxanthine, phosphatidylethanolamine 34:2p, phosphatidylethanolamine 36:1p, and pyroglutamic acid. In this embodiment, the above-mentioned plasma metabolic marker composition shows significant differences across different cancer types, enabling the differentiation of different cancer types in cancer patients and achieving pan-cancer detection.

[0061] In some embodiments, the plasma metabolic marker composition further includes at least one selected from allantoic acid, phosphatidylethanolamine 36:3p, arabinose, oxalate, and phosphatidylcholine 34:2e. That is, the plasma metabolic marker composition includes at least one selected from glycyl-L-glutamate, phosphatidylcholine 38:7, pyrrole-2-carboxylic acid, phosphatidylethanolamine 36:2p, and hypoxanthine, phosphatidylethanolamine 34:2p, phosphatidylethanolamine 36:1p, and pyroglutamate, as well as at least one selected from allantoic acid, phosphatidylethanolamine 36:3p, arabinose, oxalate, and phosphatidylcholine 34:2e. The above-mentioned plasma metabolic marker composition shows significant differences across different cancer types, enabling the differentiation of different cancer types in cancer patients and achieving pan-cancer detection.

[0062] In some embodiments, the plasma metabolic marker composition further includes at least one of L-malic acid, nicotinic acid, oxalic acid, phosphatidylethanolamine 38:5p, and uridine. That is, in this embodiment, the plasma metabolic marker composition includes at least one of glycyl-L-glutamate, phosphatidylcholine 38:7, pyrrole-2-carboxylic acid, phosphatidylethanolamine 36:2p, and at least one of hypoxanthine, phosphatidylethanolamine 34:2p, phosphatidylethanolamine 36:1p, and pyroglutamate; at least one of allantoic acid, phosphatidylethanolamine 36:3p, arabinose, oxalic acid, and phosphatidylcholine 34:2e; and at least one of L-malic acid, nicotinic acid, oxalic acid, phosphatidylethanolamine 38:5p, and uridine. The above-mentioned plasma metabolic marker composition shows significant differences across different cancer types, enabling the differentiation of different cancer types in cancer patients and achieving pan-cancer detection.

[0063] In some embodiments, the plasma metabolic marker composition comprises glycyl-L-glutamic acid, phosphatidylcholine 38:7, pyrrole-2-carboxylic acid, phosphatidylethanolamine 36:2p, hypoxanthine, phosphatidylethanolamine 34:2p, phosphatidylethanolamine 36:1p, pyroglutamic acid, allantoic acid, phosphatidylethanolamine 36:3p, arabinose, oxalic acid, and phosphatidylcholine 34:2e. The above-mentioned plasma metabolic marker composition shows significant differences across different cancer types, enabling the differentiation of different cancer types in cancer patients and achieving pan-cancer detection.

[0064] A second aspect of the present invention provides a method for screening the plasma metabolic marker composition of the present invention as described above, comprising the following steps:

[0065] S1. Extract plasma from patients with gastric cancer, colorectal cancer, and lung cancer respectively to obtain extracts;

[0066] S2. The organic phase and aqueous phase in the extract were detected using a chromatography-mass spectrometry system to obtain detection data;

[0067] S3. The detection data is processed, and then the metabolites are identified to obtain metabolomics data.

[0068] S4. Analyze the metabolomics data and screen to obtain plasma metabolic biomarker compositions.

[0069] This invention, based on metabolomics, enables high-throughput detection of small molecule metabolites in the body by preparing plasma samples. Combined with multivariate statistical analysis, it screens for significantly different metabolic biomarkers, enabling the differentiation of different cancer types in cancer patients and providing a new option for pan-cancer screening.

[0070] In step S4, in some embodiments, the step of analyzing the metabolomics data and screening to obtain the plasma metabolic biomarker composition specifically includes:

[0071] The metabolomics data were subjected to linear correlation analysis, and the Pearson correlation coefficient was calculated. The plasma metabolic biomarker composition was obtained by screening with the absolute value of the Pearson correlation coefficient being greater than a preset value.

[0072] In some implementations, a screening criterion is used where the absolute value of the Pearson correlation coefficient is greater than or equal to 0.3. The Pearson correlation coefficient is widely used to measure the degree of correlation between two variables (in this invention, the two variables refer to the intensity of metabolites and different cancer types). The Pearson correlation coefficient (i.e., the R value) ranges between -1 and 1. If the Pearson correlation coefficient is close to -1, it indicates a high negative correlation between the two variables; if the Pearson correlation coefficient is close to 1, it indicates a high positive correlation between the two variables; if the Pearson correlation coefficient is close to 0, it indicates that the two variables are independent and have no correlation. Therefore, this invention uses a screening criterion where the absolute value of the Pearson correlation coefficient is greater than or equal to 0.3, i.e., |R|≥0.3, to screen for metabolic biomarkers that are highly correlated with distinguishing different cancer types.

[0073] This invention also provides the application of the plasma metabolic marker composition described above in the present invention in the preparation of products for detecting pan-cancer types. The products include, but are not limited to, kits.

[0074] In some embodiments, the pan-cancer category includes gastric cancer, colorectal cancer, and lung cancer. In this invention, the above-mentioned plasma metabolic marker composition can be used to differentiate between the three major cancer categories of gastric cancer, colorectal cancer, and lung cancer.

[0075] The present invention will be further described below through specific embodiments.

[0076] Example 1: Detection, identification, and screening of small molecule metabolites and metabolic biomarkers in plasma samples

[0077] 1. Subject information and sample collection

[0078] The inclusion and exclusion criteria for patients with gastric cancer, lung cancer, and colorectal cancer are as follows:

[0079] Inclusion criteria: (1) All subjects obtained written informed consent before participating in the study; (2) Male or female aged ≥18 years; (3) Patients diagnosed with primary gastric cancer, lung cancer or colorectal cancer by biopsy / postoperative pathology or by comprehensive clinical assessment by a clinician; those who meet all three criteria are included.

[0080] Exclusion criteria: (1) Pregnancy or lactation; (2) Emergency or requiring resuscitation; (3) History of malignant tumors or having undergone any anti-tumor treatment before sampling; (4) Simultaneous co-occurrence of multiple primary malignant tumors; if any one of the four criteria is met, the patient will be excluded.

[0081] Plasma samples were collected from 222 patients with gastric cancer (GC), 232 patients with colorectal cancer (CRC), and 579 patients with lung cancer (LC) (Table 1). All plasma samples were collected in the morning on an empty stomach and stored at -80°C. Plasma samples from gastric cancer patients were designated as the GC group, those from colorectal cancer patients as the CRC group, and those from lung cancer patients as the LC group.

[0082] Plasma samples from the GC group (222 samples), the CRC group (232 samples), and the LC group (579 samples) were randomly divided into modeling and validation groups. The plasma samples in the modeling and validation groups were different from those in the validation groups. Specifically,

[0083] Plasma sample count in the modeling groups: 167 plasma samples in the GC group, 174 plasma samples in the CRC group, and 434 plasma samples in the LC group;

[0084] The number of plasma samples in the validation groups were: 55 plasma samples in the GC group, 58 plasma samples in the CRC group, and 145 plasma samples in the LC group (see Table 1 for details).

[0085] Table 1. Subject Information

[0086] GC group CRC group LC Group Number of samples in the modeling group (number of samples) 167 174 434 Number of samples in the validation group 55 58 145 Total (number) 222 232 579

[0087] 2. Reagents

[0088] Methanol, acetonitrile, water, acetic acid, isopropanol and methyl tert-butyl ether of mass spectrometry grade, and formic acid and ammonium acetate of chromatographic (HPLC) grade were purchased from Sigma-Aldrich, USA.

[0089] 3. Sample preparation

[0090] Take 100 μL of plasma and place it in 1000 μL of pre-cooled mixed solution (composed of methyl tert-butyl ether and methanol, with a volume ratio of methyl tert-butyl ether to methanol of 3:1). Vortex to mix and obtain the sample extract.

[0091] Add 500 μL of a mixed solution of methanol and water (methanol to water volume ratio of 3:1) to the sample extract, sonicate, let stand, vortex and centrifuge to separate the layers;

[0092] After the sample was separated into layers, 500 μL of the upper organic phase was transferred to a centrifuge tube. After the organic phase was dried, 200 μL of a mixed solution of acetonitrile and isopropanol (acetonitrile to isopropanol volume ratio 3:1) was added, and the mixture was incubated at room temperature for 15 minutes. After incubation, the centrifuge tube was vortexed and sonicated for 5 minutes, and then centrifuged at room temperature (12000 rpm) for 5 minutes. 180 μL of the supernatant was transferred from the centrifuge tube to a 2 mL glass vial, which contained the organic phase, and detected by LC-MS (liquid chromatography-mass spectrometry).

[0093] After the sample was separated into layers, 400 μL of the lower aqueous phase was transferred to a centrifuge tube, and 1100 μL of ice-cold methanol was added to precipitate the protein. The centrifuge tube was then centrifuged, and 1000 μL of the supernatant was transferred to a new centrifuge tube and dried overnight. 200 μL of water was added to the dried centrifuge tube and incubated at room temperature for 15 minutes. After incubation, the centrifuge tube was vortexed and sonicated for 5 minutes, and then centrifuged at room temperature (12000 rpm) for 5 minutes. 180 μL of the supernatant was transferred from the centrifuge tube to a 2 mL glass vial. This was the aqueous phase and was analyzed by LC-MS.

[0094] 4. Detection of small molecule metabolites:

[0095] Organic phase using Waters ACQUTTY BEH C8 1.7μm 2.1mm×100mm column, with Waters ACQUTTY used for the aqueous phase. HSS T3 1.8μm 2.1mm × 100mm column was used for small molecule separation; ACQUITY UPLC I-Class liquid chromatography system (Waters) and Q-Exactive mass spectrometry system (Thermo Fisher Scientific) were used for both liquid chromatography and mass spectrometry.

[0096] The mobile phase parameters are as follows:

[0097] Organic phases: Mobile phase A is an aqueous solution containing 0.1% acetic acid and 1% ammonium acetate (acetic acid mass content is 0.1%, ammonium acetate mass content is 1%); Mobile phase B is a mixed solution of acetonitrile and isopropanol containing 0.1% acetic acid and 1% ammonium acetate (acetonitrile to isopropanol volume ratio is 7:3, acetic acid mass content is 0.1%, ammonium acetate mass content is 1%). The separation elution gradient is as follows: 0-12 minutes 55%-89% mobile phase B, 12-19.5 minutes 100% mobile phase B;

[0098] Aqueous phase: Mobile phase A is an aqueous solution containing 0.1% formic acid (formic acid mass content is 0.1%); mobile phase B is an acetonitrile solution containing 0.1% formic acid (formic acid mass content is 0.1%). The separation elution gradient is as follows: 0-13 minutes 1%-70% mobile phase B, 13-18 minutes 99% mobile phase B;

[0099] The mass spectrometry parameters are as follows:

[0100] Mass spectrometry data were acquired in Full MS and Full MS / dd-MS2 modes (each including both positive and negative modes). The parameters used for the QExactive mass spectrometry system are as follows:

[0101] In Full MS mode, the resolution is 70,000 m / z, the scan range is 100-1500 m / z (m / z is the ratio of ion mass to charge), and the AGC (Automatic Gain Control) is 3E+6 (i.e., 3×10⁻⁶). 6 The Maximum Injection Time (IT) is 200 milliseconds.

[0102] In Full MS / dd-MS2 mode, the resolution of the secondary mass spectrometer is 17,500 m / z, the quadrupole window is 1.5 m / z, and the AGC is 1E+5 (i.e., 1×10⁻⁵). 5 The maximum ion implantation time is 50 milliseconds, and the relative collision energy of HCD (high-energy collision dissociation) is 30 eV.

[0103] 5. Metabolomics Data Processing

[0104] First, effective peak signals are extracted from all mass spectrometry data. Then, baseline correction is applied to remove noise and retain the original signal peaks. The original data is converted into central discrete data. Next, the retention times in the chromatogram corresponding to each mass spectrometry peak in the sample are corrected, and peaks with differences between samples within the error range are defined as the same peak, thus obtaining a matrix dataset. Due to unavoidable systematic fluctuations during the experiment (such as instrument fluctuations, fluctuations caused by human operation, etc.), a normalization autoencoder (NormAE) is used for homogenization to make the data distribution more reflective of the true differences between samples.

[0105] 6. Identification of metabolites

[0106] We used public databases such as the Human Metabolites Database (HMDB; www.hmdb.ca), the Metabolomics Database (Metlin Database; https: / / metlin.scripps.edu), and the Mass Spectrometry Database (http: / / www.massbank.jp / ), as well as primary and secondary chromatograms and mass spectra of standards separated under the same chromatographic column; under the conditions that the retention time difference was within 0.1 min and the mass-to-charge ratio was less than 10 ppm, we matched and identified the samples with the database and the standards.

[0107] 7. Metabolomics data analysis

[0108] Linear correlation analysis was performed on the metabolomics data of the modeling group samples in the GC, CRC, and LC groups, and the Pearson correlation coefficient (R value) was calculated. Metabolic biomarkers with high correlation to different cancer types were screened by using |R|≥0.3, ultimately obtaining 18 differentially expressed metabolites (as shown in Table 2). Among them, the linear correlation curves of hypoxanthine in the three cancer modeling group samples are shown in Table 2. Figure 1 As shown (it reflects the trend of xanthine intensity changes among the three cancers), Figures 2 to 18 (Similar in meaning), the linear correlation curves of glycyl-L-glutamate in the samples of the three cancer modeling groups are as follows: Figure 2 As shown, the linear correlation curves of allantoin in the samples of the three cancer modeling groups are as follows: Figure 3 As shown, the linear correlation curves of phosphatidylethanolamine 36:3p in the three cancer modeling groups are as follows: Figure 4 As shown, the linear correlation curves of arabinose in the three cancer modeling groups are as follows: Figure 5 As shown, the linear correlation curves of phosphatidylethanolamine 34:2p in the three cancer modeling groups are as follows: Figure 6 As shown, the linear correlation curves of L-malic acid in the samples of the three cancer modeling groups are as follows: Figure 7 As shown, the linear correlation curves of oxalic acid in the three cancer modeling groups are as follows: Figure 8 As shown, the linear correlation curves of phosphatidylethanolamine 36:1p in the three cancer modeling groups are as follows: Figure 9 As shown, the linear correlation curves of niacin in the samples of the three cancer modeling groups are as follows: Figure 10 As shown, the linear correlation curves of phosphatidylcholine 38:7 in the three cancer modeling groups are as follows: Figure 11 As shown, the linear correlation curves of pyroglutamate in the samples of the three cancer modeling groups are as follows: Figure 12 As shown, the linear correlation curves of oxalic acid in the three cancer modeling groups are as follows: Figure 13 As shown, the linear correlation curves of phosphatidylethanolamine 38:5p in the three cancer modeling groups are as follows: Figure 14 As shown, the linear correlation curves of uridine in the samples of the three cancer modeling groups are as follows: Figure 15 As shown, the linear correlation curves of phosphatidylcholine 34:2e in the three cancer modeling groups are as follows: Figure 16 As shown, the linear correlation curves of pyrrole-2-carboxylic acid in the modeling group samples of the three cancers are as follows: Figure 17 As shown, the linear correlation curves of phosphatidylethanolamine 36:2p in the three cancer modeling groups are as follows: Figure 18 As shown.

[0109] Table 2. 18 metabolic markers used to distinguish different types of cancer

[0110]

[0111]

[0112] Example 2 uses 18 metabolic biomarkers to construct a diagnostic model that distinguishes different cancer types.

[0113] To evaluate the effectiveness of the metabolic biomarkers screened in Example 1 in distinguishing different cancer types, this example again performed multivariate multi-class ROC (Receiver Operating Characteristic) analysis on these 18 metabolic biomarkers. Three-quarters of the modeling group sample data (obtained from Example 1) was randomly allocated as the training set and one-quarter as the test set. The training set was used to build and train the machine learning classification model, and the test set was used to evaluate the discriminative ability of the trained model. A Support Vector Machine (SVM) algorithm was used for 1000 random iterations. An evaluation model was constructed by statistically analyzing the average accuracy of the final model. The results showed that the AUC = 0.859 (e.g., ...). Figure 19 As shown in the figure, this demonstrates that the diagnostic model is efficient.

[0114] ROC curves are a method for studying the relationship between model sensitivity and specificity. Sensitivity is plotted on the ordinate and 1-specificity on the abscissa. The evaluation is based on the area under the curve (AUC). When the AUC is greater than 0.5, the closer the AUC is to 1, the better the model performance and the better the diagnostic effect. If it is less than 0.5, it indicates that the model's accuracy is poor.

[0115] To further verify the effectiveness of the constructed diagnostic model, plasma samples from the validation group in Example 1 were used as unknown samples. The validation group data were then input into the constructed diagnostic model for verification. The results showed that AUC = 0.844 (e.g., ...). Figure 20 As shown in the figure, this demonstrates that the diagnostic model has clinical diagnostic significance.

[0116] Example 3 uses 13 metabolic biomarkers to construct a diagnostic model that distinguishes different cancer types.

[0117] The only difference from the method in Example 2 is that when constructing the diagnostic model using the SVM method, 13 metabolic markers are used in combination: hypoxanthine, glycyl-L-glutamate, allantoic acid, phosphatidylethanolamine 36:3p, arabinose, phosphatidylethanolamine 34:2p, oxalic acid, phosphatidylethanolamine 36:1p, phosphatidylcholine 38:7, pyroglutamate, phosphatidylcholine 34:2e, pyrrole-2-carboxylic acid, and phosphatidylethanolamine 36:2p.

[0118] The results showed that AUC = 0.869, which is of clinical diagnostic significance.

[0119] To further verify the effectiveness of the constructed diagnostic model, the plasma samples from the validation group in Example 1 were used as unknown samples and placed into the constructed diagnostic model for validation. The results showed that AUC = 0.845, indicating that the diagnostic model has clinical diagnostic significance.

[0120] Example 4 uses eight metabolic biomarkers to construct a diagnostic model that distinguishes different cancer types.

[0121] The only difference from the method in Example 2 is that when constructing the diagnostic model using the SVM method, eight metabolic markers are used in combination: hypoxanthine, glycyl-L-glutamate, phosphatidylethanolamine 34:2p, phosphatidylethanolamine 36:1p, phosphatidylcholine 38:7, pyroglutamate, pyrrole-2-carboxylic acid, and phosphatidylethanolamine 36:2p.

[0122] The results showed that AUC = 0.851, which is of clinical diagnostic significance.

[0123] To further verify the effectiveness of the constructed diagnostic model, the plasma samples from the validation group in Example 1 were used as unknown samples and placed into the constructed diagnostic model for validation. The results showed that AUC = 0.844, indicating that the diagnostic model has clinical diagnostic significance.

[0124] Example 5 uses four metabolic biomarkers to construct a diagnostic model that distinguishes different cancer types.

[0125] The only difference from the method in Example 2 is that when constructing the diagnostic model using the SVM method, four metabolic markers—glycyl-L-glutamate, phosphatidylcholine 38:7, pyrrole-2-carboxylic acid, and phosphatidylethanolamine 36:2p—were combined to construct the diagnostic model. The result showed an AUC of 0.843, indicating clinical diagnostic significance.

[0126] To further verify the effectiveness of the constructed diagnostic model, the plasma samples from the validation group in Example 1 were used as unknown samples and placed into the constructed diagnostic model for validation. The results showed that AUC = 0.828, indicating that the diagnostic model has clinical diagnostic significance.

[0127] In summary, the metabolic biomarker composition of this invention exhibits significant differences across different cancer types, enabling the differentiation of cancer patients across different cancer types and achieving pan-cancer detection. Pan-cancer detection using the metabolic biomarker composition is non-invasive and minimally invasive, making it more suitable for large-scale screening and detection in the general population. Furthermore, this metabolic biomarker composition possesses high sensitivity and specificity, making it more suitable for early cancer screening, aiding in early diagnosis and treatment, and significantly improving the survival and cure rates of cancer patients.

[0128] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A plasma metabolic marker composition, characterized in that, The plasma metabolic marker composition comprises glycyl-L-glutamate, phosphatidylcholine 38:7, pyrrole-2-carboxylic acid, and phosphatidylethanolamine 36:2p.

2. The plasma metabolic marker composition according to claim 1, characterized in that, The plasma metabolic marker composition further includes at least one of hypoxanthine, phosphatidylethanolamine 34:2p, phosphatidylethanolamine 36:1p, and pyroglutamic acid.

3. The plasma metabolic marker composition according to claim 2, characterized in that, The plasma metabolic marker composition further includes at least one of allantoic acid, phosphatidylethanolamine 36:3p, arabinose, oxalic acid, and phosphatidylcholine 34:2e.

4. The plasma metabolic marker composition according to claim 3, characterized in that, The plasma metabolic marker composition further includes at least one of L-malic acid, nicotinic acid, oxalic acid, phosphatidylethanolamine 38:5p, and uridine.

5. The plasma metabolic marker composition according to claim 3, characterized in that, The plasma metabolic marker composition comprises glycyl-L-glutamic acid, phosphatidylcholine 38:7, pyrrole-2-carboxylic acid, phosphatidylethanolamine 36:2p, hypoxanthine, phosphatidylethanolamine 34:2p, phosphatidylethanolamine 36:1p, pyroglutamic acid, allantoic acid, phosphatidylethanolamine 36:3p, arabinose, oxalic acid, and phosphatidylcholine 34:2e.

6. A method for screening the plasma metabolic marker composition according to any one of claims 1-5, characterized in that, Includes the following steps: Plasma from patients with gastric cancer, colorectal cancer, and lung cancer was extracted to obtain extracts; The organic and aqueous phases in the extract were detected using chromatography-mass spectrometry to obtain detection data; The detection data is processed and then the metabolites are identified to obtain metabolomics data. The metabolomics data were analyzed to screen and obtain the plasma metabolic biomarker composition.

7. The screening method according to claim 6, characterized in that, The step of analyzing the metabolomics data and screening to obtain the plasma metabolic biomarker composition specifically includes: The metabolomics data were subjected to linear correlation analysis, and the Pearson correlation coefficient was calculated. The plasma metabolic biomarker composition was obtained by screening with the absolute value of the Pearson correlation coefficient being greater than a preset value.

8. The screening method according to claim 7, characterized in that, The absolute value of the Pearson correlation coefficient is greater than or equal to 0.3 as the screening criterion.

9. The use of a plasma metabolic marker composition according to any one of claims 1-5 in the preparation of a product for detecting pan-cancer species; wherein the pan-cancer species are gastric cancer, colorectal cancer, or lung cancer.

Images