Application of metabolic marker in preparation of product for detecting or diagnosing adverse cardiovascular event of coronary heart disease and construction method of model

Abstract

The invention relates to the technical field of metabolic markers, in particular to application of metabolic markers in preparation of products for detecting or diagnosing coronary heart disease adverse cardiovascular events and a construction method of a coronary heart disease adverse cardiovascular event diagnosis model. The coronary heart disease adverse cardiovascular event product is used for detecting or diagnosing coronary heart disease cardiovascular death events and / or coronary heart disease heart failure events. According to the application, the risks of death events and heart failure events of the coronary heart disease patients are predicted by using the specific differential metabolites, dense monitoring can be performed on the coronary heart disease patients, and effective customized treatment guidance is provided for clinicians.

Description

Technical Field

[0001] This application relates to the technical field of metabolic biomarkers, and in particular to the application of metabolic biomarkers in the preparation of products for detecting or diagnosing adverse cardiovascular events in coronary heart disease and methods for constructing diagnostic models for adverse cardiovascular events in coronary heart disease. Background Technology

[0002] Coronary artery disease (CAD) is currently the leading cause of death and morbidity worldwide; cardiovascular events, including cardiovascular death, heart failure (HF), myocardial infarction (MI), and stroke, continue to occur, imposing a significant socioeconomic burden and negatively impacting patients' quality of life. Accurate assessment of the risk of cardiovascular complications and implementation of risk-oriented secondary prevention are urgently needed. However, risk assessment for overall cardiovascular events is unreliable because the progression and occurrence of these complications, which originate from CAD, can be highly heterogeneous. Elucidating shared and unique metabolic disorders holds the potential to identify biomarkers for tailoring future cardiovascular event risk assessments, while simultaneously improving our understanding of the molecular processes involved in different outcomes.

[0003] Several circulating biomarkers, such as cardiac troponin T / I (cTn T / I) and N-terminal probrain natriuretic peptide (NT-proBNP), have been identified as predictors of adverse cardiovascular events. However, these established biomarkers are often detected after myocardial injury or significant cardiac dysfunction. For patients with cardiovascular disease (CAD), earlier detection of plasma biomarkers and identification of molecules particularly associated with cardiovascular events are crucial for prognostic assessment and timely intervention. Previous studies have reported several metabolites that can predict the risk of cardiovascular events, but these studies have underestimated specific biomarkers across different cardiovascular outcomes. Therefore, it is necessary to identify the metabolic profile of shared and unique metabolic disorders associated with cardiovascular events.

[0004] Metabolomics is a promising and powerful approach for elucidating metabolic disorders associated with pathological states. Compared to genomics and proteomics, metabolomics is the omics layer most relevant to phenotype. Given the significant heterogeneity of individual cardiovascular event mechanisms initiating from CAD, the rich information in metabolomics data has not been fully considered in cardiovascular event risk prediction. Targetless metabolomics can comprehensively detect and relatively quantify metabolites, providing a comprehensive understanding of metabolic changes during disease progression and offering a promising approach for the widespread discovery of circulating metabolite biomarkers in CAD patients.

[0005] Therefore, this application provides the application of metabolic biomarkers in the preparation of products for detecting or diagnosing adverse cardiovascular events in coronary heart disease and a method for constructing a diagnostic model for adverse cardiovascular events in coronary heart disease. Summary of the Invention

[0006] The purpose of this application is to provide the application of metabolic biomarkers in the preparation of products for detecting or diagnosing adverse cardiovascular events in coronary heart disease and a method for constructing a diagnostic model for coronary heart disease. The application described in this application uses specific differential metabolites to predict the risk of death and heart failure events in patients with coronary heart disease, enabling intensive monitoring of patients with coronary heart disease and providing clinicians with guidance for effective customized treatment.

[0007] To solve the above-mentioned technical problems, this application is implemented as follows:

[0008] The application of metabolic biomarkers in the preparation of products for detecting or diagnosing adverse cardiovascular events in coronary heart disease, wherein the products are used to detect or diagnose cardiovascular death events and / or heart failure events in coronary heart disease;

[0009] The metabolic biomarkers used for detecting or diagnosing cardiovascular death events due to coronary heart disease include one or more of the metabolic biomarkers described below:

[0010] Tetradecanoylcarnitine (Car(14:2)), Glutamine-Carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), Oxyoctanoylcarnitine (Car(8:1-O)), Homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4);

[0011] The metabolic biomarkers specifically used for detecting or diagnosing coronary heart disease and cardiovascular death events can be any one of the above-mentioned metabolic biomarkers, or any two, three, four, five, six, seven, or all eight biomarkers.

[0012] The metabolic biomarkers used for detecting or diagnosing coronary heart failure events include one or more of the following metabolic biomarkers:

[0013] Oxycarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), cortisol.

[0014] The metabolic biomarkers specifically used for detecting or diagnosing coronary heart disease and heart failure events can be any one of the above-mentioned metabolic biomarkers, or any two, three, four, five, six, or all seven biomarkers.

[0015] Furthermore, the metabolic biomarkers used to detect or diagnose cardiovascular death events caused by coronary heart disease are combined with N-terminal probrain natriuretic peptide (NT-proBNP) for diagnosis.

[0016] Furthermore, the metabolic biomarkers used to detect or diagnose cardiovascular death events caused by coronary heart disease are combined with N-terminal probrain natriuretic peptide (NT-proBNP) for diagnosis.

[0017] Furthermore, the product for adverse cardiovascular events related to coronary heart disease is a chip, test strip, or reagent kit for the detection or diagnosis of cardiovascular death events and / or heart failure events related to coronary heart disease.

[0018] Furthermore, the aforementioned product for detecting adverse cardiovascular events related to coronary heart disease is used to test biochemical samples. Specifically, the biochemical samples are blood or plasma.

[0019] On the other hand, this application provides a method for constructing a diagnostic model for adverse cardiovascular events in coronary heart disease, the method comprising the following steps:

[0020] S1. Collect the subject's blood or plasma and detect the concentration of blood metabolites;

[0021] S2. Conduct clinical diagnosis on the subjects;

[0022] S3. Based on the detection results of step S1 and the clinical diagnosis of step S2, construct a diagnostic model;

[0023] Blood metabolites of coronary artery disease-related cardiovascular death events include one or more metabolic markers as described below:

[0024] Tetradecanoylcarnitine (Car(14:2)), Glutamine-Carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), Oxyoctanoylcarnitine (Car(8:1-O)), Homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4);

[0025] Blood metabolites used to detect or diagnose coronary heart disease and heart failure events include one or more metabolic biomarkers as described below:

[0026] Oxycarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), cortisol.

[0027] Furthermore, the methods for constructing diagnostic models include, but are not limited to, classification and logistic regression, k-nearest neighbor algorithm, Naive Bayes, support vector machine, decision tree, random forest, regression tree, gradient boosting decision tree, xgboost, lightweight gradient boosting machine, gradient boosting machine, LASSO, and convolutional neural network.

[0028] Furthermore, the method used to detect the concentration of blood metabolites includes:

[0029] The method employs one or more of the following techniques: nuclear magnetic resonance spectroscopy, mass spectrometry, chromatography, ion mobility spectrometry, electrochemical detection, Raman spectroscopy, or radiolabeling. Specifically, in the specific embodiments of this application, liquid chromatography-mass spectrometry (LC-MS) is used for metabolomics analysis, and its usage and detection principle are well known in the art.

[0030] The beneficial effects of this application are as follows:

[0031] The application described in this application provides, for the first time, comprehensive evidence of metabolic abnormalities in patients with coronary heart disease during various cardiovascular events, and identifies specific metabolic biomarkers and pathways.

[0032] The application described in this application utilizes specific differential metabolites to predict the risk of death and heart failure events in patients with coronary artery disease, enabling intensive monitoring of these patients and providing clinicians with guidance for effective customized treatment. Attached Figure Description

[0033] Figure 1 This is a flowchart of non-targeted metabolomics research.

[0034] Figure 2 Inclusion and exclusion criteria for the study subjects;

[0035] Figure 3 Definition of cardiovascular events;

[0036] Figure 4 Partial least squares discriminant analysis (PLS-DA) 3D score plots for the composite cardiovascular event group and the control group;

[0037] Figure 5 Volcano plot showing the differences in 492 metabolites between composite cardiovascular events and the control group;

[0038] Figure 6 Heatmap of differential metabolites associated with complex cardiovascular events;

[0039] Figure 7 To include metabolite information that predicts coronary heart disease mortality and coronary heart failure events;

[0040] Figure 8Prediction results of individual metabolites in mortality events, as well as prediction / diagnostic models and factor coefficients;

[0041] Figure 9 The prediction results of two metabolites in the mortality event, as well as the prediction / diagnostic model and factor coefficients;

[0042] Figure 10a Predictions for three metabolites in mortality events, along with predictive / diagnostic models and factor coefficients;

[0043] Figure 10b Predictions for three metabolites in mortality events, along with predictive / diagnostic models and factor coefficients;

[0044] Figure 11a Predictions for four metabolites in mortality events, along with predictive / diagnostic models and factor coefficients;

[0045] Figure 11b Predictions for four metabolites in mortality events, along with predictive / diagnostic models and factor coefficients;

[0046] Figure 11c Predictions for four metabolites in mortality events, along with predictive / diagnostic models and factor coefficients;

[0047] Figure 12a Predictions for five metabolites in mortality events, along with predictive / diagnostic models and factor coefficients;

[0048] Figure 12b Predictions for five metabolites in mortality events, along with predictive / diagnostic models and factor coefficients;

[0049] Figure 13 Predictions for six metabolites in mortality events, along with predictive / diagnostic models and factor coefficients;

[0050] Figure 14 Predictions for seven metabolites in mortality events, along with predictive / diagnostic models and factor coefficients;

[0051] Figure 15 Predictions for eight metabolites in mortality events, along with predictive / diagnostic models and factor coefficients;

[0052] Figure 16 For single metabolites combined with NT-proBNP and predictive / diagnostic models and factor coefficients in mortality events;

[0053] Figure 17 The combined NT-proBNP and predictive / diagnostic models and factor coefficients for two metabolites in mortality events;

[0054] Figure 18aThe combined NT-proBNP and predictive / diagnostic models and factor coefficients for three metabolites in mortality events;

[0055] Figure 18b The combined NT-proBNP and predictive / diagnostic models and factor coefficients for three metabolites in mortality events;

[0056] Figure 19a The combined NT-proBNP and predictive / diagnostic models and factor coefficients for four metabolites in mortality events;

[0057] Figure 19b The combined NT-proBNP and predictive / diagnostic models and factor coefficients for four metabolites in mortality events;

[0058] Figure 19c The combined NT-proBNP and predictive / diagnostic models and factor coefficients for four metabolites in mortality events;

[0059] Figure 20 Five metabolites combined with NT-proBNP in mortality events, along with predictive / diagnostic models and factor coefficients;

[0060] Figure 21 The combined NT-proBNP model and factor coefficients for six metabolites in mortality events;

[0061] Figure 22 Seven metabolites combined with NT-proBNP in mortality events, along with predictive / diagnostic models and factor coefficients;

[0062] Figure 23 Eight metabolites combined with NT-proBNP in mortality events, along with predictive / diagnostic models and factor coefficients;

[0063] Figure 24 For the prediction results of individual metabolites in heart failure events, as well as the predictive / diagnostic models and factor coefficients;

[0064] Figure 25 The prediction results of two metabolites in heart failure events, as well as the predictive / diagnostic model and factor coefficients;

[0065] Figure 26a The prediction results of three metabolites in heart failure events, as well as the predictive / diagnostic model and factor coefficients;

[0066] Figure 26b The prediction results of three metabolites in heart failure events, as well as the predictive / diagnostic model and factor coefficients;

[0067] Figure 27a The prediction results of four metabolites in heart failure events, as well as the prediction / diagnostic model and factor coefficients;

[0068] Figure 27b The prediction results of four metabolites in heart failure events, as well as the prediction / diagnostic model and factor coefficients;

[0069] Figure 28 The prediction results of five metabolites in heart failure events, as well as the predictive / diagnostic models and factor coefficients;

[0070] Figure 29 Predictive results for six metabolites in heart failure events, along with predictive / diagnostic models and factor coefficients;

[0071] Figure 30 Predictions for seven metabolites in heart failure events, along with predictive / diagnostic models and factor coefficients;

[0072] Figure 31 For single metabolite combined with NT-proBNP and predictive / diagnostic models and factor coefficients in heart failure events;

[0073] Figure 32 The combined NT-proBNP metabolite and predictive / diagnostic models and factor coefficients in heart failure events;

[0074] Figure 33a The combined NT-proBNP and predictive / diagnostic models and factor coefficients for three metabolites in heart failure events;

[0075] Figure 33b The combined NT-proBNP and predictive / diagnostic models and factor coefficients for three metabolites in heart failure events;

[0076] Figure 34 The combined NT-proBNP and predictive / diagnostic models and factor coefficients for four metabolites in heart failure events;

[0077] Figure 35a Five metabolites combined with NT-proBNP in heart failure events, along with predictive / diagnostic models and factor coefficients;

[0078] Figure 35b Five metabolites combined with NT-proBNP in heart failure events, along with predictive / diagnostic models and factor coefficients;

[0079] Figure 36 A model and factor coefficients for six metabolites combined with NT-proBNP in heart failure events, including a predictive / diagnostic model.

[0080] Figure 37 Seven metabolites combined with NT-proBNP in heart failure events, along with predictive / diagnostic models and factor coefficients;

[0081] Figure 38a , Figure 38b , Figure 38c Differential metabolites associated with complex cardiovascular events;

[0082] Figure 39a , Figure 39b , Figure 39c , Figure 39d These are differentially expressed metabolites associated with heart failure. Detailed Implementation

[0083] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0084] The experimental research process, such as Figure 1 As shown:

[0085] 1. Subjects and Study Design

[0086] Patients participating in this study were from Qilu Hospital of Shandong University (Site 1: Jinan, China) and Zibo Central Hospital (Site 2: Zibo, China). The inclusion and exclusion criteria for patients are detailed below. Figure 2 and Figure 3 .

[0087] This study employed a nested case-control design, comprising 333 patients who experienced cardiovascular events as the case group and 333 patients who did not experience any events as the control group. The control group was randomly selected from all at-risk participants, propensity score matched 1:1 based on age, sex, body mass index, current smoking status, hypertension, diabetes, and history of myocardial infarction. To ensure the reliability and reproducibility of the selection of differentially expressed metabolites, patients were randomly assigned to the discovery and validation sets in a 1:1 ratio (334 patients in the discovery set and 332 patients in the validation set). This study has been approved by the Research Ethics Committee of Qilu Hospital, Shandong University, and accepted by Zibo Central Hospital.

[0088] 2. Cardiovascular event assessment

[0089] Cardiovascular events were defined as cardiovascular death, heart failure, and myocardial infarction / stroke. The composite index of cardiovascular events and individual cardiovascular events were defined as outcome events. Outcome events were recorded via 12-month follow-up telephone surveys conducted by trained research assistants at each participating hospital.

[0090] 3. Metabolomics based on liquid chromatography-mass spectrometry (LC-MS)

[0091] Data acquisition was performed using non-targeted metabolomics based on liquid chromatography-mass spectrometry (LC-MS).

[0092] 4. Statistical Analysis

[0093] Partial least squares discriminant analysis (PLS-DA) was used to examine the differential metabolites between the cardiovascular event group and the control group.

[0094] 5. Specific sampling, testing and analysis process

[0095] Patient plasma samples (50 μL) were extracted using 150 μL of methanol and internal standards (d3-leucine and d6-phenylalanine). The samples were then vortexed for 30 seconds and treated with an ultrasonic processor for 15 minutes. To precipitate proteins, the samples were incubated at -20°C for 1 hour and then centrifuged at 13,500 rpm and 4°C for 15 minutes. The resulting supernatant was transferred to high-performance liquid chromatography (HPLC) vials and stored at -80°C prior to LC-MS / MS analysis. Frozen plasma samples were stored at -80°C at the Shandong Provincial Clinical Research Center for Emergency and Critical Care Medicine, Jinan, China.

[0096] The supernatant was analyzed by LC-MS / MS. LC separation was performed using a Waters ACQUITY UPLC BEHAmide column (1.7 μm particle size; 10 / 0 mm (length) × 2.1 mm (inner diameter)) and a Kinetex C18 column (2.6 μm, 2.1 × 100 mm), with the column temperature maintained at 25 °C. For hydrophilic interaction liquid chromatography (HILIC) analysis, mobile phase A consisted of 25 mM NH4OH + 25 mM NH4OAc dissolved in water, and mobile phase B was ACN, suitable for both positive ion mode (ESI+) and negative ion mode (ESI-). The flow rate was 0.5 mL / min, with the following gradient settings: 0-0.5 min, B at 95%; 0.5-7 min, B decreasing linearly from 95% to 65%; 7-8 min, B decreasing linearly from 65% to 40%; 8-9 min, B at 40%; 9-9.1 min, B increasing linearly from 40% to 95%; 9.1-12 min, B at 95%. The injection volume was 2 μL. For reversed-phase liquid chromatography (RPLC) analysis, mobile phase A was 0.01% acetic acid dissolved in water, and mobile phase B was a mixture of IPA and ACN (1:1), suitable for both positive ion mode (ESI+) and negative ion mode (ESI-). The flow rate was 0.3 mL / min, and the gradient settings were as follows: 0-1 min: B = 1%; 1-8 min: B = 99%; 8-9 min: B = 99%; 9.0-9.1 min: B decreases linearly from 99% to 1%; 9.1-12 min: B = 1%. The injection volume was 2 μL.

[0097] All samples were injected randomly during data acquisition. Data acquisition was performed using a Thermo Scientific Vanquish UHPLC system in conjunction with a Thermo Scientific Orbitrap Exploris 480.

[0098] Data acquisition was performed in a single sample in full MS-scan mode with switching between positive and negative ion polarities.

[0099] Quality control (QC) samples were acquired using Information-Dependent Acquisition (IDA) mode to obtain MS / MS spectra. Source parameters were set as follows: spray voltage was 3000V in positive ion mode and -3000V in negative ion mode. Assist gas heater temperature was set to 350℃. Jacket gas was set to 50 alb. Assist gas was set to 15 alb. Capillary temperature was set to 400℃. Full scan resolution in positive or negative ion mode was set to 60,000, and the AGC target was set to 1e6. Maximum integration time was set to 100 ms. Mass range was set to 70-1200 Da. For dd-MS2 settings, MS resolution was set to 30,000, and the AGC target was set to 1e5. Maximum integration time was set to 60 ms. Top N was set to 6. Isolation width was set to 1.0 m / zDa.

[0100] MS / MS spectra of QC samples were acquired at SNCE 20-30-40%. Dynamic exclusion was set to 3.0 s, and isotope exclusion was enabled.

[0101] Raw MS data (.raw) files were converted to mzXML format using ProteoWizard (version 3.0.20360). Then, we used the XCMS software package (version 3.2) to perform peak detection, retention time correction, and peak alignment on the mass spectrometry data. Key parameter settings were as follows: method "centWave"; ppm = 10; snhr = 3; peakwidth = c(5,30); minfrac = 0.5. Metabolite annotation was performed using MetDNA software. Metabolite annotation parameters were set to "HILIC" or "RP" depending on the liquid chromatography mode, and collision energy was set to "30" or "SNCE_20_30_40%".

[0102] 6. Analysis and Conclusion

[0103] 6.1 Shared metabolic profile with cardiovascular events

[0104] Compared with the control group, patients in the cardiovascular event group showed a clearly distinguishable metabolic profile. Figure 4Metabolic pathway analysis revealed disruptions in 19 metabolic pathways associated with a combination of cardiovascular events. These pathways included tyrosine metabolism, cysteine ​​and methionine metabolism, pentose and glucuronide interconversion, lysine degradation, and fatty acid biosynthesis. We identified 82 metabolites as differential metabolites predicting the combined risk of cardiovascular events. Figure 5 After adjusting for variables related to thrombolytic therapy for myocardial infarction (TIMI), NT-proBNP, and hs-cTnT, a total of 23 differentially expressed metabolites were significantly associated with the combined risk of cardiovascular events. Figure 6 ).

[0105] Of these 23 metabolites, a significant proportion were medium- and long-chain acylcarnitines (13 ≤ carbon ≤ 25). Nine acylcarnitines were significantly elevated in patients with cardiovascular events compared to the control group. Plasma levels of phthalates and 5-acetamido-6-amino-3-methyluracil (AAMU) were decreased, while other metabolites were increased. Furthermore, we identified combinations of key differentially expressed metabolites for predicting overall cardiovascular event risk, which may serve as potential biomarkers. These key metabolite combinations comprised 14 metabolites, including 12 upregulated and 2 downregulated metabolites; among which… Figure 38a , Figure 38b , Figure 38c For differentially expressed metabolites associated with composite cardiovascular events, the following values ​​are provided: FDR q value from Wilcoxon rank-sum test; p-value for the association between metabolite and composite cardiovascular event after adjusting for TIM variables, hs-cTnT, and NT-proBNP; β value from LASSO algorithm; FC = fold change; FDR = false detection rate; LASSO = minimum absolute value contraction and selection operator; LPC = lysophosphatidylcholine; LPE = lysophosphatidylethanolamine; m / z = mass-to-charge ratio; RT = retention time (seconds).

[0106] Figure 39a , Figure 39b , Figure 39c , Figure 39d These are differentially expressed metabolites associated with heart failure. Among them: FDR q value of Wilcoxon rank-sum test; P-values ​​for the association between metabolites and composite cardiovascular events after adjusting for TIM variables, hs-cTnT, and NT-proBNP; The β value of the LASSO algorithm; FC = multiple change; FDR = false detection rate; LASSO = minimum absolute value shrinkage and selection operator; LPC = lysophosphatidylcholine; LPE = lysophosphatidylethanolamine; m / z = mass-to-charge ratio; RT = retention time (seconds);

[0107] 6.2 Screening for target metabolites for cardiovascular death / heart failure using genetic algorithms

[0108] Based on training and validation set data, targeted metabolites for measuring cardiovascular death / heart failure were screened from 492 metabolites. The training set was split in a 7:3 ratio, and a genetic algorithm was used to screen the metabolite set from 70% of the training set population. The remaining 30% of the training and validation set data were used for validation. Figure 7 As shown, a total of 6 targeted measurement metabolites were screened (panel1);

[0109] Among them, the cardiovascular death metabolites (AUC = 0.687) include: homoarginine, 3-methoxy-4-hydroxyphenylglycol sulfate, glutarylcarnitine, and vanillylmandelic acid.

[0110] Heart failure metabolites (AUC = 0.861) include: N-acetyl-arginine, oleoylcarnitine (Car(18:1)), homoarginine, and 3-methoxy-4-hydroxyphenyl glycol sulfate.

[0111] 6.3 In addition to the above 6 metabolites, further increase the number of targeted metabolites to be measured.

[0112] 6.3.1 Metabolite Screening for Targeted Measurement of Cardiovascular Mortality

[0113] The following criteria were used to screen targeted metabolites for discovery:

[0114] ①False discovery rate(FDR)q value<0.05and|log1.2(Fold change)|>1;

[0115] ② After adjusting for clinical risk factors, it remained significantly associated with cardiovascular death (P<0.05). A total of 14 metabolites were screened. Based on this, the Lasso method was used to find the optimal set of metabolites, and finally 9 metabolites associated with cardiovascular death were selected. Based on the external validation AUC of the prediction model after combining a single metabolite with 4 targeted measurement metabolites and NT-proBNP and the feasibility of actual measurement, the following 3 metabolites were selected as supplementary metabolites for the second round of screening (panel 2 for CVdeath):

[0116] Tetradecadienoylcarnitine (Car(14:2)) , N-[3-(2-oxopyrrolidin-1-yl)propyl]acetamide , 3-Hydroxyoctanoic acid.

[0117] 6.3.2 Screening of Metabolites for Targeted Measurement in Heart Failure

[0118] The following criteria were used to screen targeted metabolites for discovery:

[0119] ①False discovery rate(FDR)q value<0.05and|log1.2(Fold change)|>1;

[0120] ② After adjusting for clinical risk factors, it still showed a significant association with heart failure (P<0.05). A total of 21 metabolites were screened. Based on this, the Lasso method was used to find the optimal set of metabolites, and finally 13 heart failure metabolites were screened. Based on the external validation AUC of the predictive model after combining a single metabolite with 4 targeted measurement metabolites and NT-proBNP and the feasibility of actual measurement, the following 3 metabolites were selected as supplementary differential metabolites for the second round of screening (panel 2 for heart failure):

[0121] N4-Acetylcytidine, LPC(O-16:0), 11beta-hydroxyandrost-4-ene-3,17-dione.

[0122] Further review and discussion revealed low signal intensities in the metabolites vanillylmandelic acid and 11beta-hydroxyandrost-4-ene-3,17-dione. Based on correlation and predictive model performance, N,N-Dimethylarginine (ADMA) and Cortisol were selected as alternative metabolites, respectively. Considering that Oxooctanoylcarnitine (Car(8:1-O)) can serve as an alternative metabolite for vanillylmandelic acid and 11beta-hydroxyandrost-4-ene-3,17-dione in cardiovascular death and heart failure outcomes, it was also included in the targeted measurement metabolite range.

[0123] Ultimately, eight metabolites were included to predict coronary heart disease mortality events, and seven metabolites were included to predict coronary heart disease heart failure events. Details of the metabolites are as follows ( Figure 7 ):

[0124] Eight metabolites used to predict coronary heart disease mortality events are: tetradecenoic carnitine (Car(14:2)), glutamine carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoic carnitine (Car(8:1-O)), homoarginine (HArg), and 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4).

[0125] The seven metabolites used to predict coronary heart failure events are: oxoctylcarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), and cortisol.

[0126] 7. Predictive / Diagnostic Models and Factor Coefficients

[0127] Taking cardiovascular death as an example, this paper demonstrates the detailed process of predictive model construction and external validation in the screening of targeted metabolites:

[0128] ① First, the training set (n=334) and validation set (n=332) data were extracted. The data frame contained 8 metabolites and cardiovascular mortality outcomes. ② Based on the glm function, a logistic prediction model was constructed on the training set, with the 8 candidate differentially expressed metabolites as predictors and cardiovascular mortality outcomes as predictors. ③ The performance of the prediction model was evaluated on the validation set data using the predict function. ④ The same steps were followed to incorporate NT-proBNP and construct a prediction model.

[0129] Car(8:1-O) is oxoctylcarnitine, HArg is homoarginine, MHPG-SO4 is 3-methoxy-4-hydroxyphenylethylene glycol sulfate, Car(18:1) is oleocarnitine, Ac-Arg is N-acetyl-arginine, ac4C is N4-acetylcytidine, and Cortisol is cortisol.

[0130] Validation was performed on the training and validation sets, and the results were as follows: Figures 8-37 Predictive / diagnostic results: Among them Figure 8 The prediction results of individual metabolites in mortality events, as well as the predictive / diagnostic models and factor coefficients. Figure 9The prediction results for two metabolites in mortality events, as well as the predictive / diagnostic model and factor coefficients. Figure 10a The prediction results of three metabolites in mortality events, as well as the predictive / diagnostic model and factor coefficients. Figure 10b The prediction results of three metabolites in mortality events, as well as the predictive / diagnostic model and factor coefficients. Figure 11a The prediction results for four metabolites in the mortality event, as well as the predictive / diagnostic model and factor coefficients, are presented. Figure 11b The prediction results for four metabolites in the mortality event, as well as the predictive / diagnostic model and factor coefficients, are presented. Figure 11c The prediction results for four metabolites in the mortality event, as well as the predictive / diagnostic model and factor coefficients, are presented. Figure 12a The prediction results for five metabolites in mortality events, along with the predictive / diagnostic model and factor coefficients. Figure 12b The prediction results for five metabolites in mortality events, along with the predictive / diagnostic model and factor coefficients. Figure 13 The prediction results for six metabolites in mortality events, along with the predictive / diagnostic model and factor coefficients. Figure 14 The prediction results for seven metabolites in mortality events, along with the predictive / diagnostic model and factor coefficients. Figure 15 The prediction results for eight metabolites in mortality events, along with the predictive / diagnostic model and factor coefficients. Figure 16 For single metabolites combined with NT-proBNP and predictive / diagnostic models and factor coefficients in mortality events, Figure 17 The study included two metabolites combined with NT-proBNP in mortality events, along with predictive / diagnostic models and factor coefficients. Figure 18a The study included three metabolites combined with NT-proBNP in mortality events, along with predictive / diagnostic models and factor coefficients. Figure 18b The study included three metabolites combined with NT-proBNP in mortality events, along with predictive / diagnostic models and factor coefficients. Figure 19a The study included four metabolites combined with NT-proBNP, along with predictive / diagnostic models and factor coefficients in mortality events. Figure 19b The study included four metabolites combined with NT-proBNP, along with predictive / diagnostic models and factor coefficients in mortality events. Figure 19c The study included four metabolites combined with NT-proBNP, along with predictive / diagnostic models and factor coefficients in mortality events. Figure 20 The study included five metabolites combined with NT-proBNP, along with predictive / diagnostic models and factor coefficients in mortality events. Figure 21 The study included six metabolites combined with NT-proBNP, along with predictive / diagnostic models and factor coefficients in mortality events. Figure 22 The study included seven metabolites combined with NT-proBNP, along with predictive / diagnostic models and factor coefficients in mortality events. Figure 23 The model and factor coefficients for eight metabolites combined with NT-proBNP in mortality events were analyzed.

[0131] AUC (Area Under Curve) represents the area under the curve and is used to indicate prediction accuracy. A higher AUC value, meaning a larger area under the curve, indicates higher prediction accuracy. NT-proBNP is N-terminal probrain natriuretic peptide, Car(14:2) is tetradecenoic acid carnitine, Glu-Car is glutamine carnitine, NOPA is N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide, 3HA is 3-hydroxyoctanoic acid, ADMA is N,N-dimethylarginine, Car(8:1-O) is oxoctanoic acid carnitine, HArg is homoarginine, and MHPG-SO4 is 3-methoxy-4-hydroxyphenylethylene glycol sulfate.

[0132] Therefore, the metabolic biomarkers used for detecting or diagnosing cardiovascular death events due to coronary heart disease include one or more of the following metabolic biomarkers:

[0133] Tetradecanoylcarnitine (Car(14:2)), Glutamine-Carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), Oxyoctanoylcarnitine (Car(8:1-O)), Homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4);

[0134] Furthermore, such as Figure 8 As shown, metabolic markers can be any one of the following: tetradecenoic carnitine (Car(14:2)), glutamine carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoic acid carnitine (Car(8:1-O)), homoarginine (HArg), or 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4). Specific predictive / diagnostic models and factor coefficients are as follows: Figure 8 As shown, for example, the prediction / diagnosis model and factor coefficients are:

[0135] Logit(P)=-0.705+0.865*Car(14:2)、

[0136] Logit(P)=-0.733+0.56*Glu-Car、

[0137] Logit(P)=-0.737+0.674*NOPA、

[0138] Logit(P)=-0.724+0.517*3HA、

[0139] Logit(P)=-0.732+0.825*ADMA、

[0140] Logit(P)=-0.676+0.866*Car(8:1-O)、

[0141] Logit(P)=-0.794-0.901*HArg、

[0142] Logit(P) = -0.689 + 1.038 * MHPG-SO4, any one of them.

[0143] Furthermore, such as Figure 9 As shown, the combination of metabolic markers can be any two of the following: tetradecenoic carnitine (Car(14:2)), glutamine carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoic carnitine (Car(8:1-O)), homoarginine (HArg), and 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4); for example, it can be a combination of tetradecenoic carnitine (Car(14:2)) & glutamine carnitine (Glu-Car), or a combination of N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA) & 3-hydroxyoctanoic acid (3HA). Specific predictive / diagnostic models and factor coefficients, such as... Figure 9 As shown, for example, the prediction / diagnosis model and factor coefficients are Logit(P) = -0.706 + 0.749 * Car(14:2) + 0.238 * Glu-Car.

[0144] Logit(P)=-0.711+0.709*Car(14:2)+0.471*NOPA、

[0145] Logit(P)=-0.713+0.767*Car(14:2)+0.187*3HA、

[0146] Logit(P)=-0.706+0.632*Car(14:2)+0.581*ADMA、

[0147] Logit(P)=-0.68+0.615*Car(14:2)+0.405*Car(8:1-O)、

[0148] Logit(P)=-0.757+0.657*Car(14:2)-0.687*HArg、

[0149] Logit(P)=-0.684+0.383*Car(14:2)+0.774*MHPG-SO4、

[0150] Logit(P)=-0.739+0.342*Glu-Car+0.536*NOPA、

[0151] Logit(P) = any one of -0.742 + 0.492 * Glu - Car + 0.431 * 3HA.

[0152] Furthermore, such as Figure 10a , 10b As shown, the combination of metabolic markers can be any three of the following: tetradecenoic carnitine (Car(14:2)), glutamine carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoic carnitine (Car(8:1-O)), homoarginine (HArg), and 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4); for example, it can be a combination of tetradecenoic carnitine (Car(14:2)) & glutamine carnitine (Glu-Car) & N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA) & 3-hydroxyoctanoic acid (3HA), or a combination of N,N-dimethylarginine (ADMA) & oxoctanoic carnitine (Car(8:1-O)) & homoarginine (HArg). Specific prediction / diagnostic models and factor coefficients, such as Figure 10a , 10b As shown, for example, predictive / diagnostic models and factor coefficients are

[0153] Logit(P)=-0.711+0.668*Car(14:2)+0.1*Glu-Car+0.445*NOPA、

[0154] Logit(P)=-0.716+0.627*Car(14:2)+0.258*Glu-Car+0.211*3HA、

[0155] Logit(P) = any one of -0.706 + 0.596 * Car(14:2) + 0.094 * Glu-Car + 0.554 * ADMA.

[0156] Furthermore, such as Figure 11a , 11bAs shown in 11c, the combination of metabolic markers is either tetradecenoic carnitine (Car(14:2)), glutamine carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoic acid carnitine (Car(8:1-O)), homoarginine (HArg), or 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4). Any four of the following; for example, it could be a combination of tetradecenoic carnitine (Car(14:2)) & glutamine carnitine (Glu-Car) & N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA) & 3-hydroxyoctanoic acid (3HA), or a combination of N,N-dimethylarginine (ADMA) & oxoctanoic carnitine (Car(8:1-O)) & homoarginine (HArg) & 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4). Specific predictive / diagnostic models and factor coefficients, such as... Figure 11a , 11b As shown in 11c, for example, the predictive / diagnostic model and factor coefficients are

[0157] Logit(P)=-0.725+0.517*Car(14:2)+0.114*Glu-Car+0.464*NOPA+0.252*3HA、

[0158] Logit(P)=-0.708+0.592*Car(14:2)+0.05*Glu-Car+0.274*NOPA+0.395*ADMA、

[0159] Logit(P) = any one of -0.699 + 0.566*Car(14:2) + 0.038*Glu-Car + 0.43*NOPA + 0.218*Car(8:1-O).

[0160] Furthermore, such as Figure 13As shown, the combination of metabolic markers can be any five of the following: tetradecenoic carnitine (Car(14:2)), glutamine carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoic carnitine (Car(8:1-O)), homoarginine (HArg), and 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4); for example, it can be tetradecenoic carnitine (Car(14:2)), glutamine carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoic carnitine (Car(8:1-O)), homoarginine (HArg), and 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4). Combinations of carnitine (Car(14:2)) & glutamine carnitine (Glu-Car) & N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA) & 3-hydroxyoctanoic acid (3HA) & N,N-dimethylarginine (ADMA), and combinations of 3-hydroxyoctanoic acid (3HA) & N,N-dimethylarginine (ADMA) & oxoctanoic acid carnitine (Car(8:1-O)) & homoarginine (HArg) & 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4). Specific predictive / diagnostic models and factor coefficients, such as... Figure 12a , 12b For example, the predictive / diagnostic model and factor coefficients are Logit(P) = -0.748 + 0.357*Car(14:2) + 0.187*Glu-Car + 0.162*3HA + 0.202*Car(8:1-O) - 0.659*HArg.

[0161] Logit(P)=-0.748+0.357*Car(14:2)+0.187*Glu-Car+0.162*3HA+0.202*Car(8:1-O)-0.659*HArg,

[0162] Logit(P)=-0.689+0.247*Car(14:2)-0.009*Glu-Car+0.175*3HA+0.101*Car(8:1-O)+0.739*MHPG-SO4,

[0163] Logit(P) = any one of -0.739 + 0.217 * Car (14:2) + 0.078 * Glu-Car + 0.14 * 3HA - 0.586 * HArg + 0.543 * MHPG-SO4.

[0164] Furthermore, such as Figure 13As shown, the combination of metabolic markers can be any six of the following: tetradecenoic carnitine (Car(14:2)), glutamine carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoic carnitine (Car(8:1-O)), homoarginine (HArg), and 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4); for example, it can be tetradecenoic carnitine (Car(14:2)) & glutamine carnitine (Glu-Carnitine) The combination of N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), and oxooctanoylcarnitine (Car(8:1-O)), and the combination of N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxooctanoylcarnitine (Car(8:1-O)), homoarginine (HArg), and 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4). Specific predictive / diagnostic models and factor coefficients, such as... Figure 13 As shown, for example, predictive / diagnostic models and factor coefficients are

[0165] Logit(P)=-0.738-0.001*Glu-Car+0.091*NOPA+0.355*ADMA+0.177*Car(8:1-O)-0.6*HArg+0.373*MHPG-SO4,

[0166] Logit(P)=-0.748+0.028*Glu-Car+0.174*3HA+0.397*ADMA+0.128*Car(8:1-O)-0.597*HArg+0.35*MHPG-SO4,

[0167] Logit(P) = any one of -0.746 + 0.136 * NOPA + 0.194 * 3HA + 0.329 * ADMA + 0.133 * Car(8:1-O) - 0.571 * HArg + 0.332 * MHPG-SO4.

[0168] Furthermore, such as Figure 14As shown, the combination of metabolic markers can be any seven of the following: tetradecenoic carnitine (Car(14:2)), glutamine carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoic acid carnitine (Car(8:1-O)), homoarginine (HArg), and 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4); for example, it can be tetradecenoic carnitine (Car(14:2)) & glutamine carnitine (Glu-Car) & N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoic acid carnitine (Car(8:1-O)), homoarginine (HArg), and 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4). Combinations of pyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoylcarnitine (Car(8:1-O)), and homoarginine (HArg), and combinations of glutamine carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoylcarnitine (Car(8:1-O)), homoarginine (HArg), and 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4). Specific predictive / diagnostic models and factor coefficients, such as... Figure 14 As shown, for example, the prediction / diagnosis model and factor coefficients are Logit(P) = -0.744 + 0.21*Car(14:2) + 0.019*Glu-Car + 0.135*3HA + 0.398*ADMA + 0.051*Car(8:1-O) - 0.588*HArg + 0.288*MHPG-SO4.

[0169] Logit(P)=-0.743+0.223*Car(14:2)+0.146*NOPA+0.154*3HA+0.324*ADMA+0.046*Car(8:1-O)-0.563*HArg+0.259*MHPG-SO4,

[0170] Logit(P) = any one of -0.747 + 0.014 * Glu-Car + 0.135 * NOPA + 0.195 * 3HA + 0.328 * ADMA + 0.127 * Car(8:1-O) - 0.573 * HArg + 0.327 * MHPG-SO4.

[0171] Furthermore, such as Figure 15The combination of metabolic markers shown is tetradecenoic carnitine (Car(14:2)), glutamine carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoic carnitine (Car(8:1-O)), homoarginine (HArg), and 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4). Specific predictive / diagnostic models and factor coefficients are as follows... Figure 15 As shown, for example, predictive / diagnostic models and factor coefficients are

[0172] Logit(P)=-0.743+0.223*Car(14:2)+0.004*Glu-Car+0.145*NOPA+0.154*3HA+0.324*ADMA+0.044*Car(8:1-O)-0.564*HArg+0.257*MHPG-SO4.

[0173] Similarly, such as Figure 16 , Figure 17 , Figure 18a , Figure 18b , Figure 19a , Figure 19b , Figure 19c , Figure 20 , Figure 21 , Figure 22 , Figure 23 As shown, a combination of metabolic biomarkers, in conjunction with N-terminal pro-brain natriuretic peptide (NT-proBNP), is used to predict or diagnose cardiovascular death events in patients with coronary heart disease. This combination of metabolic biomarkers includes one or more of the following:

[0174] Tetradecanoylcarnitine (Car(14:2)), glutamine carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), oxoctanoylcarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4). Specific predictive / diagnostic models and factor coefficients, such as... Figure 16 , Figure 17 , Figure 18a , Figure 18b , Figure 19a , Figure 19b , Figure 19c , Figure 20 , Figure 21 , Figure 22 , Figure 23As shown; for example, the predictive / diagnostic model and factor coefficients of a single metabolic marker combined with N-terminal pro-brain natriuretic peptide: Logit(P)=-5.791+0.762*NT-proBNP+0.38*Car(14:2),

[0175] Logit(P)=-6.269+0.83*NT-proBNP+0.177*Glu-Car、

[0176] Logit(P)=-6.018+0.794*NT-proBNP+0.382*NOPA、

[0177] Logit(P)=-6.276+0.831*NT-proBNP+0.369*3HA、

[0178] Logit(P)=-5.876+0.771*NT-proBNP+0.382*ADMA、

[0179] Logit(P)=-5.879+0.778*NT-proBNP+0.465*Car(8:1-O)、

[0180] Logit(P)=-5.812+0.759*NT-proBNP-0.561*HArg、

[0181] Logit(P) = -5.585 + 0.732 * NT - proBNP + 0.461 * MHPG - SO4, any one of these values;

[0182] For example, the predictive / diagnostic model and factor coefficients for the combination of two metabolic markers and N-terminal probrain natriuretic peptide are as follows:

[0183] Logit(P)=-5.756+0.756*NT-proBNP+0.353*Car(14:2)+0.058*Glu-Car、

[0184] Either Logit(P) = -5.572 + 0.729 * NT - proBNP + 0.279 * Car(14:2) + 0.317 * NOPA or Logit(P) = -5.886 + 0.775 * NT - proBNP + 0.223 * Car(14:2) + 0.278 * 3HA.

[0185] For example, the predictive / diagnostic model and factor coefficients for the combination of three metabolic biomarkers and N-terminal probrain natriuretic peptide are as follows:

[0186] Logit(P)=-5.589+0.731*NT-proBNP+0.292*Car(14:2)-0.033*Glu-Car+0.325*NOPA、

[0187] Logit(P)=-5.841+0.768*NT-proBNP+0.182*Car(14:2)+0.078*Glu-Car+0.285*3HA、

[0188] Logit(P) = -5.529 + 0.721 * NT-proBNP + 0.263 * Car(14:2) - 0.006 * Glu-Car + 0.295 * ADMA (any one of these).

[0189] For example, the predictive / diagnostic model and factor coefficients for the combination of four metabolic biomarkers and N-terminal probrain natriuretic peptide are as follows:

[0190] Logit(P)=-5.668+0.741*NT-proBNP+0.101*Car(14:2)-0.022*Glu-Car+0.348*NOPA+0.319*3HA,

[0191] Logit(P)=-5.511+0.719*NT-proBNP+0.261*Car(14:2)-0.046*Glu-Car+0.254*NOPA+0.143*ADMA,

[0192] Logit(P) = any one of -5.607 + 0.736 * NT - proBNP + 0.157 * Car (14:2) - 0.09 * Glu - Car + 0.294 * NOPA + 0.253 * Car (8:1 - O).

[0193] For example, the predictive / diagnostic model and factor coefficients for the combination of five metabolic biomarkers and N-terminal probrain natriuretic peptide are as follows:

[0194] Logit(P)=-5.609+0.732*NT-proBNP+0.084*Car(14:2)-0.033*Glu-Car+0.294*NOPA+0.31*3HA+0.111*ADMA,

[0195] Logit(P)=-5.693+0.747*NT-proBNP-0.033*Car(14:2)-0.078*Glu-Car+0.32*NOPA+0.316*3HA+0.252*Car(8:1-O),

[0196] Logit(P) = any one of -5.375 + 0.694 * NT - proBNP + 0.044 * Car(14:2) - 0.021 * Glu - Car + 0.28 * NOPA + 0.281 * 3HA - 0.434 * HArg.

[0197] For example, the predictive / diagnostic model and factor coefficients for the combination of six metabolic biomarkers and N-terminal probrain natriuretic peptide are as follows:

[0198] Logit(P)=-5.642+0.739*NT-proBNP-0.038*Car(14:2)-0.084*Glu-Car+0.279*NOPA+0.309*3HA+0.089*ADMA+0.234*Car(8:1-O),

[0199] Logit(P)=-5.305+0.683*NT-proBNP+0.024*Car(14:2)-0.031*Glu-Car+0.22*NOPA+0.271*3HA+0.119*ADMA-0.436*HArg,

[0200] Logit(P) = any one of -5.537 + 0.722 * NT-proBNP + 0.015 * Car(14:2) - 0.074 * Glu-Car + 0.273 * NOPA + 0.302 * 3HA + 0.071 * ADMA + 0.187 * MHPG-SO4.

[0201] For example, the predictive / diagnostic model and factor coefficients for a combination of seven metabolic biomarkers plus N-terminal probrain natriuretic peptide are as follows:

[0202] Logit(P)=-5.279+0.68*NT-proBNP-0.071*Car(14:2)-0.045*Glu-Car+0.24*3HA+0.206*ADMA+0.157*Car(8:1-O)-0.449*HArg+0.072*MHPG-SO4,

[0203] Logit(P)=-5.304+0.684*NT-proBNP-0.057*Car(14:2)+0.201*NOPA+0.273*3HA+0.1*ADMA+0.121*Car(8:1-O)-0.427*HArg-0.002*MHPG-SO4,

[0204] Logit(P) = any one of -5.294 + 0.683 * NT-proBNP - 0.065 * Glu-Car + 0.21 * NOPA + 0.257 * 3HA + 0.099 * ADMA + 0.118 * Car(8:1-O) - 0.422 * HArg + 0.008 * MHPG-SO4.

[0205] For example, the predictive / diagnostic model and factor coefficients for a combination of eight metabolic biomarkers plus N-terminal probrain natriuretic peptide are as follows:

[0206] Logit(P)=-5.326+0.688*NT-proBNP-0.056*Car(14:2)-0.065*Glu-Car+0.208*NOPA+0.269*3HA+0.099*ADMA+0.144*Car(8:1-O)-0.422*HArg+0.023*MHPG-SO4.

[0207] Figure 24 This study presents prediction results for individual metabolites in heart failure events, along with predictive / diagnostic models and factor coefficients. Figure 25 This study presents prediction results for two metabolites in heart failure events, along with predictive / diagnostic models and factor coefficients. Figure 26a The prediction results of three metabolites in heart failure events, as well as the predictive / diagnostic model and factor coefficients. Figure 26b The prediction results of three metabolites in heart failure events, as well as the predictive / diagnostic model and factor coefficients. Figure 27a The prediction results of four metabolites in heart failure events, as well as the predictive / diagnostic model and factor coefficients. Figure 27b The prediction results of four metabolites in heart failure events, as well as the predictive / diagnostic model and factor coefficients. Figure 28 The prediction results of five metabolites in heart failure events, as well as the predictive / diagnostic models and factor coefficients. Figure 29 The prediction results of six metabolites in heart failure events, as well as the predictive / diagnostic models and factor coefficients. Figure 30 Predictive results for seven metabolites in heart failure events, along with predictive / diagnostic models and factor coefficients. Figure 31 The study investigated single metabolite combined with NT-proBNP and predictive / diagnostic models and factor coefficients in heart failure events. Figure 32 The study investigated the combination of two metabolites, NT-proBNP, and predictive / diagnostic models and factor coefficients in heart failure events. Figure 33a The study included three metabolites combined with NT-proBNP in heart failure events, along with predictive / diagnostic models and factor coefficients. Figure 33b The study included three metabolites combined with NT-proBNP in heart failure events, along with predictive / diagnostic models and factor coefficients. Figure 34The study included four metabolites combined with NT-proBNP, along with predictive / diagnostic models and factor coefficients for heart failure events. Figure 35a Five metabolites combined with NT-proBNP, along with predictive / diagnostic models and factor coefficients, were used to analyze heart failure events. Figure 35b Five metabolites combined with NT-proBNP, along with predictive / diagnostic models and factor coefficients, were used to analyze heart failure events. Figure 36 A model and factor coefficients for six metabolites combined with NT-proBNP in predictive and diagnostic models of heart failure events. Figure 37 This section presents a model and factor coefficients for seven metabolites combined with NT-proBNP in predictive / diagnostic models of heart failure events. AUC (Area Under Curve), representing the area under the curve, indicates predictive accuracy; a higher AUC value, i.e., a larger area under the curve, indicates higher predictive accuracy. Car(8:1-O) is oxoctylcarnitine, HArg is homoarginine, MHPG-SO4 is 3-methoxy-4-hydroxyphenylethylene glycol sulfate, Car(18:1) is oleocarnitine, Ac-Arg is N-acetyl-arginine, ac4C is N4-acetylcytidine, and Cortisol is cortisol.

[0208] Therefore, the metabolic biomarkers used for detecting or diagnosing coronary heart disease and heart failure events include one or more metabolic biomarkers as described below:

[0209] Oxycarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), cortisol.

[0210] Furthermore, such as Figure 24 The metabolic biomarkers shown can be any one of the following: oxoctylcarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), or cortisol; Specific predictive / diagnostic models and factor coefficients, such as... Figure 24 As shown, for example, the prediction / diagnosis model and factor coefficients are Logit(P) = -1.221 + 0.725 * Car(8:1 - O).

[0211] Logit(P)=-1.473-1.083*HArg、

[0212] Logit(P)=-1.217+0.96*MHPG-SO4、

[0213] Logit(P)=-1.288+0.545*Car(18:1)、

[0214] Logit(P)=-1.377-0.665*Ac-Arg、

[0215] Logit(P)=-1.246+0.939*ac4C、

[0216] Logit(P) = -1.361 + 0.697 * Cortisol, any one of them.

[0217] Furthermore, such as Figure 25 The combination of metabolic markers shown can be any two of the following: oxoctylcarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), and cortisol. For example, it can be a combination of oxoctylcarnitine (Car(8:1-O)) & homoarginine (HArg), or a combination of 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4) & oleocarnitine (Car(18:1)). Specific predictive / diagnostic models and factor coefficients, such as... Figure 25 As shown, for example, the prediction / diagnosis model and factor coefficients are Logit(P) = -1.412 + 0.43 * Car(8:1-O) - 0.93 * HArg.

[0218] Logit(P)=-1.214+0.209*Car(8:1-O)+0.809*MHPG-SO4、

[0219] Logit(P)=-1.247+0.621*Car(8:1-O)+0.407*Car(18:1)、

[0220] Logit(P) = any one of -1.343 + 0.81 * Car(8:1-O) - 0.715 * Ac - Arg.

[0221] Furthermore, such as Figure 26a , Figure 26bThe combination of metabolic markers shown can be any three of the following: oxoctylcarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), and cortisol. For example, it could be a combination of oxoctylcarnitine (Car(8:1-O)) & homoarginine (HArg) & 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), or a combination of oleocarnitine (Car(18:1)) & N-acetyl-arginine (Ac-Arg) & N4-acetylcytidine (ac4C). Specific predictive / diagnostic models and factor coefficients, such as... Figure 26a , Figure 26b As shown, for example, predictive / diagnostic models and factor coefficients are

[0222] Logit(P)=-1.375+0.154*Car(8:1-O)-0.815*HArg+0.466*MHPG-SO4、

[0223] Logit(P)=-1.425+0.326*Car(8:1-O)-0.92*HArg+0.37*Car(18:1)、

[0224] Logit(P)=-1.45+0.537*Car(8:1-O)-0.756*HArg-0.473*Ac-Arg,

[0225] Logit(P) = any one of -1.389 + 0.129 * Car(8:1-O) - 0.812 * HArg + 0.562 * ac4C.

[0226] Furthermore, such as Figure 27a , Figure 28The combination of metabolic markers shown can be any four of the following: oxoctylcarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), and cortisol. For example, it could be a combination of oxoctylcarnitine (Car(8:1-O)) & homoarginine (HArg) & 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4) & oleocarnitine (Car(18:1)), or a combination of homoarginine (HArg) & 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4) & oleocarnitine (Car(18:1)) & N-acetyl-arginine (Ac-Arg). Specific predictive / diagnostic models and factor coefficients, such as... Figure 27a , Figure 28 As shown, for example, predictive / diagnostic models and factor coefficients are

[0227] Logit(P)=-1.396+0.115*Car(8:1-O)-0.83*HArg+0.374*MHPG-SO4+0.325*Car(18:1)、

[0228] Logit(P)=-1.422+0.122*Car(8:1-O)-0.567*HArg+0.71*MHPG-SO4-0.59*Ac-Arg,

[0229] Logit(P)=-1.371+0.073*Car(8:1-O)-0.78*HArg+0.205*MHPG-SO4+0.471*ac4C、

[0230] Logit(P) = any one of -1.422 + 0.162 * Car(8:1-O) - 0.769 * HArg + 0.403 * MHPG-SO4 + 0.488 * Cortisol.

[0231] Furthermore, such as Figure 28The combination of metabolic markers shown can be any five of the following: oxoctylcarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), and cortisol. For example, it could be a combination of oxoctylcarnitine (Car(8:1-O)), homoarginine (HArg), and 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), or a combination of oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), and N4-acetylcytidine (ac4C). Specific predictive / diagnostic models and factor coefficients, such as... Figure 28 As shown, for example, the prediction / diagnosis model and factor coefficients are Logit(P) = -1.457 + 0.083*Car(8:1-O) - 0.573*HArg + 0.606*MHPG-SO4 + 0.362*Car(18:1) - 0.621*Ac-Arg.

[0232] Logit(P)=-1.398+0.035*Car(8:1-O)-0.795*HArg+0.116*MHPG-SO4+0.317*Car(18:1)+0.464*ac4C,

[0233] Logit(P) = any one of -1.44 + 0.128 * Car(8:1-O) - 0.785 * HArg + 0.328 * MHPG-SO4 + 0.279 * Car(18:1) + 0.461 * Cortisol.

[0234] Furthermore, such as Figure 29The combination of metabolic markers shown can be any six of the following: oxoctylcarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleoylcarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), and cortisol. For example, it could be oxoctylcarnitine (Car(8:1-O)) & homoarginine (HArg) & 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4). The combinations of 4-hydroxyphenylene glycol sulfate (MHPG-SO4), oleoylcarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), and N4-acetylcytidine (ac4C), and homoarginine (HArg), 3-methoxy-4-hydroxyphenylene glycol sulfate (MHPG-SO4), oleoylcarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), and cortisol. Specific predictive / diagnostic models and factor coefficients, such as... Figure 29 As shown, for example, predictive / diagnostic models and factor coefficients are

[0235] Logit(P)=-1.464-0.002*Car(8:1-O)-0.527*HArg+0.318*MHPG-SO4+0.354*Car(18:1)-0.643*Ac-Arg+0.512*ac4C,

[0236] Logit(P)=-1.476+0.115*Car(8:1-O)-0.545*HArg+0.544*MHPG-SO4+0.311*Car(18:1)-0.565*Ac-Arg+0.41*Cortisol,

[0237] Logit(P)=-1.457+0.027*Car(8:1-O)-0.754*HArg-0.008*MHPG-SO4+0.252*Car(18:1)+0.586*ac4C+0.513*Cortisol,

[0238] Logit(P)=-1.465+0.03*Car(8:1-O)-0.513*HArg+0.264*MHPG-SO4-0.543*Ac-Arg+0.612*ac4C+0.493*Cortisol,

[0239] Logit(P)=-1.509+0.065*Car(8:1-O)-0.552*HArg+0.298*Car(18:1)-0.545*Ac-Arg+0.659*ac4C+0.466*Cortisol,

[0240] Logit(P)=-1.445+0.018*Car(8:1-O)+0.46*MHPG-SO4+0.279*Car(18:1)-0.754*Ac-Arg+0.622*ac4C+0.475*Cortisol,

[0241] Logit(P) = any one of -1.494 - 0.515 * HArg + 0.203 * MHPG-SO4 + 0.288 * Car(18:1) - 0.569 * Ac-Arg + 0.59 * ac4C + 0.453 * Cortisol.

[0242] Furthermore, such as Figure 30 The combination of metabolic markers shown is either oxoctylcarnitine (Car(8:1-O)) & homoarginine (HArg) & 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4) & oleocarnitine (Car(18:1)) & N-acetyl-arginine (Ac-Arg) & N4-acetylcytidine (ac4C) & cortisol. Specific predictive / diagnostic models and factor coefficients, such as... Figure 30 As shown, the factor coefficients can be

[0243] Logit(P)=-1.494+0.007*Car(8:1-O)-0.515*HArg+0.199*MHPG-SO4+0.288*Car(18:1)-0.569*Ac-Arg+0.588*ac4C+0.453*Cortisol.

[0244] Similarly, Figure 31 , Figure 32 , Figure 33a , Figure 33b , Figure 34 , Figure 35a , Figure 35b , Figure 36 , Figure 37 A combination of metabolic biomarkers, in conjunction with N-terminal pro-brain natriuretic peptide (NT-proBNP), is used to predict or diagnose coronary heart failure events. This combination of metabolic biomarkers includes one or more of the following:

[0245] Oxycarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), cortisol.

[0246] For example, the predictive / diagnostic model and factor coefficients for a single metabolic biomarker combined with N-terminal probrain natriuretic peptide are: Logit(P) = -7.677 + 0.947 * NT-proBNP + 0.348 * Car(8:1-O).

[0247] Logit(P)=-7.231+0.863*NT-proBNP-0.639*HArg、

[0248] Logit(P)=-7.41+0.906*NT-proBNP+0.388*MHPG-SO4、

[0249] Logit(P)=-7.97+0.982*NT-proBNP+0.216*Car(18:1)、

[0250] Logit(P)=-8.295+1.016*NT-proBNP-0.549*Ac-Arg、

[0251] Logit(P)=-7.353+0.897*NT-proBNP+0.469*ac4C、

[0252] Logit(P) = any one of -8.015 + 0.983 * NT - proBNP + 0.465 * Cortisol.

[0253] For example, the predictive / diagnostic model and factor coefficients for the combination of two metabolic biomarkers and N-terminal probrain natriuretic peptide are as follows:

[0254] Logit(P)=-6.965+0.828*NT-proBNP+0.212*Car(8:1-O)-0.598*HArgLogit(P)=-7.358+0.899*NT-proBNP+0.169*Car(8:1-O)+0.277*MHPG-SO4,

[0255] Logit(P)=-7.495+0.919*NT-proBNP+0.313*Car(8:1-O)+0.172*Car(18:1)、

[0256] Logit(P)=-7.538+0.911*NT-proBNP+0.454*Car(8:1-O)-0.608*Ac-Arg、

[0257] Logit(P)=-7.277+0.887*NT-proBNP+0.12*Car(8:1-O)+0.401*ac4C、Logit(P)=-7.472+0.909*NT-proBNP+0.329*Car(8:1-O)+0.461*Cortisol、

[0258] Logit(P)=-6.924+0.821*NT-proBNP-0.584*HArg+0.184*MHPG-SO4

[0259] Logit(P)=-6.927+0.817*NT-proBNP-0.645*HArg+0.237*Car(18:1)、

[0260] Logit(P)=-7.426+0.884*NT-proBNP-0.51*HArg-0.423*Ac-Arg,

[0261] Logit(P) = any one of -6.747 + 0.795 * NT - proBNP - 0.57 * HArg + 0.313 * ac4C.

[0262] For example, the predictive / diagnostic model and factor coefficients for the combination of three metabolic biomarkers and N-terminal probrain natriuretic peptide are as follows:

[0263] Logit(P)=-6.901+0.819*NT-proBNP+0.165*Car(8:1-O)-0.584*HArg+0.074*MHPG-SO4,

[0264] Logit(P)=-6.781+0.799*NT-proBNP+0.164*Car(8:1-O)-0.61*HArg+0.208*Car(18:1),

[0265] Logit(P)=-7.06+0.835*NT-proBNP+0.326*Car(8:1-O)-0.43*HArg-0.482*Ac-Arg,

[0266] Logit(P) = any one of -6.728 + 0.793 * NT - proBNP + 0.056 * Car(8:1-O) - 0.565 * HArg + 0.281 * ac4C.

[0267] For example, the predictive / diagnostic model and factor coefficients for the combination of four metabolic biomarkers and N-terminal probrain natriuretic peptide are as follows:

[0268] Logit(P)=-6.76+0.796*NT-proBNP+0.145*Car(8:1-O)-0.604*HArg+0.03*MHPG-SO4+0.205*Car(18:1),

[0269] Logit(P)=-6.877+0.807*NT-proBNP+0.157*Car(8:1-O)-0.372*HArg+0.26*MHPG-SO4-0.524*Ac-Arg,

[0270] Logit(P)=-6.775+0.799*NT-proBNP+0.088*Car(8:1-O)-0.579*HArg-0.098*MHPG-SO4+0.324*ac4C、

[0271] Logit(P) = any one of -6.808 + 0.799*NT-proBNP + 0.166*Car(8:1-O) - 0.569*HArg + 0.031*MHPG-SO4 + 0.416*Cortisol.

[0272] For example, the predictive / diagnostic model and factor coefficients for five metabolic biomarkers combined with N-terminal probrain natriuretic peptide are as follows:

[0273] ogit(P)=-6.726+0.782*NT-proBNP+0.129*Car(8:1-O)-0.386*HArg+0.209*MHPG-SO4+0.251*Car(18:1)-0.548*Ac-Arg,

[0274] Logit(P)=-6.637+0.777*NT-proBNP+0.068*Car(8:1-O)-0.598*HArg-0.134*MHPG-SO4+0.197*Car(18:1)+0.317*ac4C,

[0275] Logit(P) = any one of -6.722 + 0.784 * NT-proBNP + 0.146 * Car(8:1-O) - 0.586 * HArg - 0.009 * MHPG-SO4 + 0.183 * Car(18:1) + 0.405 * Cortisol.

[0276] For example, the predictive / diagnostic model and factor coefficients for six metabolic biomarkers combined with N-terminal probrain natriuretic peptide are as follows:

[0277] Logit(P)=-6.591+0.76*NT-proBNP+0.04*Car(8:1-O)-0.368*HArg+0.025*MHPG-SO4+0.239*Car(18:1)-0.571*Ac-Arg+0.376*ac4C,

[0278] Logit(P)=-6.735+0.779*NT-proBNP+0.125*Car(8:1-O)-0.376*HArg+0.171*MHPG-SO4+0.224*Car(18:1)-0.514*Ac-Arg+0.382*Cortisol,

[0279] Logit(P)=-6.535+0.755*NT-proBNP+0.057*Car(8:1-O)-0.58*HArg-0.22*MHPG-SO4+0.165*Car(18:1)+0.393*ac4C+0.431*Cortisol,

[0280] Logit(P)=-6.643+0.767*NT-proBNP+0.045*Car(8:1-O)-0.347*HArg-0.022*MHPG-SO4-0.514*Ac-Arg+0.452*ac4C+0.421*Cortisol,

[0281] Logit(P)=-6.551+0.751*NT-proBNP+0.008*Car(8:1-O)-0.351*HArg+0.201*Car(18:1)-0.536*Ac-Arg+0.409*ac4C+0.401*Cortisol,

[0282] Logit(P)=-6.826+0.796*NT-proBNP+0.017*Car(8:1-O)+0.087*MHPG-SO4+0.182*Car(18:1)-0.648*Ac-Arg+0.485*ac4C+0.424*Cortisol,

[0283] Logit(P) = any one of the following: -6.56 + 0.752 * NT-proBNP - 0.358 * HArg - 0.032 * MHPG-SO4 + 0.204 * Car(18:1) - 0.532 * Ac-Arg + 0.434 * ac4C + 0.404 * Cortisol

[0284] For example, the predictive / diagnostic model and factor coefficients for seven metabolic biomarkers combined with N-terminal probrain natriuretic peptide are as follows:

[0285] Logit(P)=-6.561+0.752*NT-proBNP+0.027*Car(8:1-O)-0.359*HArg-0.048*MHPG-SO4+0.203*Car(18:1)-0.53*Ac-Arg+0.427*ac4C+0.404*Cortisol.

[0286] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. The application of metabolic biomarkers in the preparation of products for detecting or diagnosing adverse cardiovascular events in coronary heart disease, characterized in that, The aforementioned product for adverse cardiovascular events related to coronary heart disease is used to detect or diagnose cardiovascular death events and / or heart failure events related to coronary heart disease. The metabolic biomarkers used for detecting or diagnosing cardiovascular death events due to coronary heart disease include one or more of the metabolic biomarkers described below: Tetradecanoylcarnitine (Car(14:2)), Glutamine-Carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), Oxyoctanoylcarnitine (Car(8:1-O)), Homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4); The metabolic biomarkers used for detecting or diagnosing coronary heart failure events include one or more of the following metabolic biomarkers: Oxycarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), cortisol.

2. The application according to claim 1, characterized in that, The aforementioned metabolic biomarkers are used to detect or diagnose cardiovascular death events caused by coronary heart disease, in combination with N-terminal probrain natriuretic peptide (NT-proBNP) for diagnosis.

3. The application according to claim 1, characterized in that, The aforementioned metabolic biomarkers are used to detect or diagnose cardiovascular death events caused by coronary heart disease, in combination with N-terminal probrain natriuretic peptide (NT-proBNP) for diagnosis.

4. The application according to claim 1, characterized in that, The aforementioned products for adverse cardiovascular events related to coronary heart disease are chips, test strips, or kits used for the detection or diagnosis of cardiovascular death events and / or heart failure events related to coronary heart disease.

5. The application according to claim 1, characterized in that, The product for detecting adverse cardiovascular events related to coronary heart disease is used to test biochemical samples.

6. A method for constructing a diagnostic model for adverse cardiovascular events in coronary heart disease, characterized in that, The construction method includes the following steps: S1. Collect the subject's blood or plasma and detect the concentration of metabolites; S2. Conduct clinical diagnosis on the subjects; S3. Based on the detection results of step S1 and the clinical diagnosis of step S2, construct a diagnostic model; Blood metabolites of coronary artery disease-related cardiovascular death events include one or more metabolic markers as described below: Tetradecanoylcarnitine (Car(14:2)), Glutamine-Carnitine (Glu-Car), N-[3-(2-oxopyrrolidone-1-yl)propyl]acetamide (NOPA), 3-hydroxyoctanoic acid (3HA), N,N-dimethylarginine (ADMA), Oxyoctanoylcarnitine (Car(8:1-O)), Homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4); Blood metabolites used to detect or diagnose coronary heart disease and heart failure events include one or more metabolic biomarkers as described below: Oxycarnitine (Car(8:1-O)), homoarginine (HArg), 3-methoxy-4-hydroxyphenylethylene glycol sulfate (MHPG-SO4), oleocarnitine (Car(18:1)), N-acetyl-arginine (Ac-Arg), N4-acetylcytidine (ac4C), cortisol.

7. The construction method according to claim 6, characterized in that, The methods for constructing diagnostic models include, but are not limited to, classification and logistic regression, k-nearest neighbor algorithm, Naive Bayes, support vector machine, decision tree, random forest, regression tree, gradient boosting decision tree, xgboost, lightweight gradient boosting machine, gradient boosting machine, LASSO, and convolutional neural network.

8. The construction method according to claim 6, characterized in that, The methods used to detect the concentration of blood metabolites include: One or more of the following techniques: nuclear magnetic resonance spectroscopy, mass spectrometry, chromatography, ion mobility spectroscopy, electrochemical detection, Raman spectroscopy, or radioactive tagging.

9. The construction method according to claim 7, characterized in that, In the constructed diagnostic model, the prediction / diagnostic prediction / diagnostic model and factor coefficients for single metabolites in coronary heart disease cardiovascular death events are as follows: Logit(P)=-0.705+0.865*Car(14:2)、 Logit(P)=-0.733+0.56*Glu-Car、 Logit(P)=-0.737+0.674*NOPA、 Logit(P)=-0.724+0.517*3HA、 Logit(P)=-0.732+0.825*ADMA、 Logit(P)=-0.676+0.866*Car(8:1-O)、 Logit(P)=-0.794-0.901*HArg、 Logit(P) = -0.689 + 1.038 * MHPG-SO4; The predictive and diagnostic model and factor coefficients for the single metabolite combined with N-terminal probrain natriuretic peptide in coronary heart disease cardiovascular mortality events are as follows: Logit(P)=-5.791+0.762*NT-proBNP+0.38*Car(14:2)、 Logit(P)=-6.269+0.83*NT-proBNP+0.177*Glu-Car、 Logit(P)=-6.018+0.794*NT-proBNP+0.382*NOPA、 Logit(P)=-6.276+0.831*NT-proBNP+0.369*3HA、 Logit(P)=-5.876+0.771*NT-proBNP+0.382*ADMA、 Logit(P)=-5.879+0.778*NT-proBNP+0.465*Car(8:1-O)、 Logit(P)=-5.812+0.759*NT-proBNP-0.561*HArg、 Logit(P) = any one of -5.585 + 0.732 * NT - proBNP + 0.461 * MHPG - SO4; The predictive / diagnostic model and factor coefficients for a single metabolite in coronary heart disease and heart failure events are as follows: Logit(P)=-1.221+0.725*Car(8:1-O)、 Logit(P)=-1.473-1.083*HArg、 Logit(P)=-1.217+0.96*MHPG-SO4、 Logit(P)=-1.288+0.545*Car(18:1)、 Logit(P)=-1.377-0.665*Ac-Arg、 Logit(P)=-1.246+0.939*ac4C、 Logit(P) = -1.361 + 0.697 * any one of Cortisol; The predictive / diagnostic model and factor coefficients for the single metabolite combined with N-terminal probrain natriuretic peptide in coronary heart disease and heart failure events are as follows: Logit(P)=-7.677+0.947*NT-proBNP+0.348*Car(8:1-O)、 Logit(P)=-7.231+0.863*NT-proBNP-0.639*HArg、 Logit(P)=-7.41+0.906*NT-proBNP+0.388*MHPG-SO4、 Logit(P)=-7.97+0.982*NT-proBNP+0.216*Car(18:1)、 Logit(P)=-8.295+1.016*NT-proBNP-0.549*Ac-Arg、 Logit(P)=-7.353+0.897*NT-proBNP+0.469*ac4C、 Logit(P) = any one of -8.015 + 0.983 * NT - proBNP + 0.465 * Cortisol.

10. The construction method according to claim 7, characterized in that, In the constructed diagnostic model, the prediction / diagnostic model and factor coefficients for the seven metabolites of coronary heart disease cardiovascular death events are as follows: Logit(P)=-0.744+0.21*Car(14:2)+0.019*Glu-Car+0.135*3HA+0.398*ADMA+0.051*Car(8:1-O)-0.588*HArg+0.288*MHPG-SO4, Logit(P)=-0.743+0.223*Car(14:2)+0.146*NOPA+0.154*3HA+0.324*ADMA+0.046*Car(8:1-O)-0.563*HArg+0.259*MHPG-SO4, Logit(P) = any one of -0.747 + 0.014 * Glu-Car + 0.135 * NOPA + 0.195 * 3HA + 0.328 * ADMA + 0.127 * Car(8:1-O) - 0.573 * HArg + 0.327 * MHPG-SO4; The predictive and diagnostic model and factor coefficients for the seven metabolites combined with N-terminal probrain natriuretic peptide in coronary heart disease cardiovascular death events are as follows: Logit(P)=-5.279+0.68*NT-proBNP-0.071*Car(14:2)-0.045*Glu-Car+0.24*3HA+0.206*ADMA+0.157*Car(8:1-O)-0.449*HArg+0.072*MHPG-SO4, Logit(P)=-5.304+0.684*NT-proBNP-0.057*Car(14:2)+0.201*NOPA+0.273*3HA+0.1*ADMA+0.121*Car(8:1-O)-0.427*HArg-0.002*MHPG-SO4, Logit(P) = any one of the following: -5.294 + 0.683 * NT-proBNP - 0.065 * Glu-Car + 0.21 * NOPA + 0.257 * 3HA + 0.099 * ADMA + 0.118 * Car(8:1-O) - 0.422 * HArg + 0.008 * MHPG-SO4; The predictive / diagnostic models and factor coefficients for the six metabolites in coronary heart disease and heart failure events are as follows: Logit(P)=-1.464-0.002*Car(8:1-O)-0.527*HArg+0.318*MHPG-SO4+0.354*Car(18:1)-0.643*Ac-Arg+0.512*ac4C, Logit(P)=-1.476+0.115*Car(8:1-O)-0.545*HArg+0.544*MHPG-SO4+0.311*Car(18:1)-0.565*Ac-Arg+0.41*Cortisol, Logit(P)=-1.457+0.027*Car(8:1-O)-0.754*HArg-0.008*MHPG-SO4+0.252*Car(18:1)+0.586*ac4C+0.513*Cortisol, Logit(P)=-1.465+0.03*Car(8:1-O)-0.513*HArg+0.264*MHPG-SO4-0.543*Ac-Arg+0.612*ac4C+0.493*Cortisol, Logit(P)=-1.509+0.065*Car(8:1-O)-0.552*HArg+0.298*Car(18:1)-0.545*Ac-Arg+0.659*ac4C+0.466*Cortisol, Logit(P)=-1.445+0.018*Car(8:1-O)+0.46*MHPG-SO4+0.279*Car(18:1)-0.754*Ac-Arg+0.622*ac4C+0.475*Cortisol, Logit(P) = any one of -1.494 - 0.515 * HArg + 0.203 * MHPG-SO4 + 0.288 * Car(18:1) - 0.569 * Ac-Arg + 0.59 * ac4C + 0.453 * Cortisol; The predictive / diagnostic model and factor coefficients for the six metabolites combined with N-terminal probrain natriuretic peptide in coronary heart disease and heart failure events are as follows: Logit(P)=-6.591+0.76*NT-proBNP+0.04*Car(8:1-O)-0.368*HArg+0.025*MHPG-SO4+0.239*Car(18:1)-0.571*Ac-Arg+0.376*ac4C, Logit(P)=-6.735+0.779*NT-proBNP+0.125*Car(8:1-O)-0.376*HArg+0.171*MHPG-SO4+0.224*Car(18:1)-0.514*Ac-Arg+0.382*Cortisol, Logit(P)=-6.535+0.755*NT-proBNP+0.057*Car(8:1-O)-0.58*HArg-0.22*MHPG-SO4+0.165*Car(18:1)+0.393*ac4C+0.431*Cortisol, Logit(P)=-6.643+0.767*NT-proBNP+0.045*Car(8:1-O)-0.347*HArg-0.022*MHPG-SO4-0.514*Ac-Arg+0.452*ac4C+0.421*Cortisol, Logit(P)=-6.551+0.751*NT-proBNP+0.008*Car(8:1-O)-0.351*HArg+0.201*Car(18:1)-0.536*Ac-Arg+0.409*ac4C+0.401*Cortisol, Logit(P)=-6.826+0.796*NT-proBNP+0.017*Car(8:1-O)+0.087*MHP G-SO4+0.182*Car(18:1)-0.648*Ac-Arg+0.485*ac4C+0.424*Cortisol, Logit(P) = any one of the following: -6.56 + 0.752 * NT-proBNP - 0.358 * HArg - 0.032 * MHPG-SO4 + 0.204 * Car(18:1) - 0.532 * Ac-Arg + 0.434 * ac4C + 0.404 * Cortisol 11. The construction method according to claim 7, characterized in that, In the constructed diagnostic model, the prediction / diagnostic model and factor coefficients for the eight metabolites of coronary heart disease cardiovascular death events are as follows: Logit(P)=-0.743+0.223*Car(14:2)+0.004*Glu-Car+0.145*NOPA+0.154*3HA+0.324*ADMA+0.044*Car(8:1-O)-0.564*HArg+0.257*MHPG-SO4; The predictive and diagnostic model and factor coefficients for the eight metabolites combined with N-terminal probrain natriuretic peptide in coronary heart disease cardiovascular mortality events are as follows: Logit(P)=-5.326+0.688*NT-proBNP-0.056*Car(14:2)-0.065*Glu-Car+0.208*NOPA+0.269*3HA+0.099*ADMA+0.144*Car(8:1-O)-0.422*HAr g+0.023*MHPG-SO4; The predictive / diagnostic models and factor coefficients for the seven metabolites in coronary heart disease and heart failure events are as follows: Logit(P)=-1.494+0.007*Car(8:1-O)-0.515*HArg+0.199*MHPG-SO4+0.288*Car(18:1)-0.569*Ac-Arg+0.588*ac4C+0.453*Cortisol; The predictive / diagnostic model and factor coefficients for the seven metabolites combined with N-terminal probrain natriuretic peptide in coronary heart disease and heart failure events are as follows: Logit(P)=-6.561+0.752*NT-proBNP+0.027*Car(8:1-O)-0.359*HArg-0.048*MHPG-SO4+0.203*Car(18:1)-0.53*Ac-Arg+0.427*ac4C+0.404*Cortisol.

Images