Screening and application of metabolic markers of rifampicin-resistant tuberculosis

Abstract

The application belongs to the technical field of biology and particularly relates to screening and application of a rifampicin-resistant tuberculosis urine metabolic marker. The urine metabolic marker comprises 20 metabolites. After verification, the level change of the 20 metabolites is indeed related to rifampicin-resistant tuberculosis. The application further provides a screening method of a rifampicin-resistant tuberculosis metabolic marker, a diagnostic rifampicin-resistant tuberculosis disease detection kit and a rifampicin-resistant tuberculosis disease diagnosis model. The 20 urine metabolic markers screened are used for rifampicin-resistant tuberculosis detection, have high accuracy and are of great significance for diagnosis of rifampicin-resistant tuberculosis.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of biotechnology, and particularly relates to screening and application of a metabolic marker of rifampicin-resistant tuberculosis. BACKGROUND

[0002] Tuberculosis (TB) is a chronic and consumptive zoonosis caused by Mycobacterium tuberculosis, and is also one of the global fatal diseases. In recent years, the application of rapid molecular detection technology has significantly improved the diagnostic efficiency of tuberculosis. Early screening and identification of drug-resistant or sensitive tuberculosis and timely diagnosis and treatment are important means to effectively control the spread of tuberculosis.

[0003] At present, the diagnosis of drug-resistant tuberculosis based on the phenotypic culture detection method has low sensitivity and long time consumption, and strict microbiological safety precautions are still needed. Although molecular biology detection technologies such as nucleic acid amplification detection technology (GeneXpert MTB / RIF, GeneXpert MTB / RIF Ultra, Xpert MTB / XDR), linear probe technology, gene sequencing technology, Truenat technology, MTB cell-free DNA (MTB-cfDNA) detection technology, and matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) detection technology have the characteristics of high efficiency, rapidness, specificity, and sensitivity, they also have some shortcomings. The results of nucleic acid amplification detection need to be comprehensively evaluated by clinicians, and false positives should be vigilant. The results of gene sequencing detection have a number of sequences with diagnostic value, and the reliability of the type of detection specimen and the clinical significance of the type of detected pathogen are not clear. The new tuberculosis detection technologies such as Truenat technology, MTB-cfDNA, and MALDI-TOF MS detection still need large-sample prospective studies to clarify their clinical significance. At the same time, the high cost of some molecular biology detection technologies limits their application and promotion in primary hospitals. Therefore, it is urgent to find a sensitive, specific, convenient, and effective early detection method, and a non-invasive sampling detection procedure is crucial for controlling rifampicin-resistant tuberculosis infection, blocking transmission, and inhibiting prevalence.

[0004] Metabolomics can provide key information on downstream products in cells and metabolic processes, and the health / disease status of specific tissues or organs. Currently, this technology has been applied in drug development, disease diagnosis, drug metabolism, adverse drug reactions, and monitoring of treatment effects. Many biological fluid samples, such as urine, blood, and tissue homogenate, can be used for metabolomics analysis. Metabolomics uses modern detection techniques combined with bioinformatics analysis methods to qualitatively or quantitatively detect the dynamic rules of the types and quantities of endogenous small molecule metabolites in tissues, cells or body fluids in the body, and obtain differential metabolic markers under the influence of different physiological and pathological stimuli or environmental factors.

[0005] Compared with blood, urine can truly reflect the body condition without the influence of balance mechanism, and has the advantages of simple collection and non-invasive sampling, and is widely used in infants and patients who cannot obtain sputum specimens. The detection of related metabolic markers in urine to determine the occurrence, development and prognosis of diseases is increasingly attracting the attention of researchers. Urine is easy to collect in large quantities and continuously non-invasively, which is helpful for metabolite identification, quantification and subsequent data analysis; at the same time, the protein content in urine is low, the processing process is relatively simple, the metabolites in urine are relatively small, thermodynamically stable, and the molecular interactions of metabolites in urine samples are less. Therefore, by studying the urine of tuberculosis patients and screening specific effective urine metabolic markers, it is crucial for the diagnosis and differential diagnosis of rifampicin-resistant tuberculosis.

[0006] In summary, how to provide more potential urine metabolic markers as a rapid, efficient, sensitive, specific and economical diagnostic method, and improve the detection efficiency is a problem that those skilled in the art need to solve. SUMMARY

[0007] In a first aspect, the present application provides a metabolic marker for differential diagnosis and / or screening of rifampicin-resistant tuberculosis, wherein the metabolic marker is one or more of Dihydroberberine, Dimethyl Sulfoxide, 1-Palmitoyl-2-Oleoyl-3-Linoleoyl-Rac-Glycerol, Choline, N,N-Dimethylarginine, Pyridoxine, D-Alanyl-D-alanine, Proline-Hydroxyproline, 2-Acetoxy-4-Pentadecylbenzoic Acid, 1-Cinnamoylpyrrolidine, 4-(Cytisin-12-Amido)-Benzoic Acid, Phosphatidylethanolamine Lyso 20, C18-Sphingosine, Erucamide, Beta-Hydroxymyristic Acid, (2-Oxo-2,3-Dihydro-1H-Indol-3-Yl)Acetic Acid, Cycloserine, 4-Pyridoxic Acid, Difluoro(Perfluoromethoxy)Acetic Acid, Beta-Alanine, with down-regulation or up-regulation of abundance.

[0008] Further, when the log2FoldChange value of the abundance of one or more of the metabolic markers Dihydroberberine, Dimethyl Sulfoxide, Choline, N,N-Dimethylarginine, Pyridoxine, D-Alanyl-D-alanine, Proline-Hydroxyproline, 4-(Cytisin-12-Amido)-Benzoic Acid, Phosphatidylethanolamine Lyso 20, (2-Oxo-2,3-Dihydro-lH-Indol-3-Yl) Acetic Acid, Cycloserine, 4-Pyridoxic Acid, Difluoro(Perfluoromethoxy) Acetic Acid, Beta-Alanine is < -0.26, i.e. the abundance is down-regulated, it indicates that the patient has a high risk of suffering from rifampicin-resistant tuberculosis.

[0009] Further, when the log2FoldChange value of the abundance of one or more of the metabolic markers 1-Palmitoyl-2-Oleoyl-3-Linoleoyl-Rac-Glycerol, 2-Acetoxy-4-Pentadecylbenzoic Acid, 1-Cinnamoylpyrrolidine, C18-Sphingosine, Erucamide, Beta-Hydroxymyristic Acid is > 0.26, i.e. the abundance is up-regulated, it indicates that the patient has a high risk of suffering from rifampicin-resistant tuberculosis.

[0010] Further, the rifampicin-resistant tuberculosis includes: rifampicin-resistant tuberculosis (RR-TB), multidrug-resistant tuberculosis (MDR-TB), pre-extensively drug-resistant tuberculosis (pre-XDR-TB) and extensive drug-resistant tuberculosis (XDR-TB).

[0011] Further, the metabolic marker is from the urine of the subject.

[0012] In a second aspect, the present application provides the use of the metabolic marker of the first aspect in the preparation of a medicament for differential diagnosis and / or screening of rifampicin-resistant tuberculosis.

[0013] Further, the use is for differential diagnosis and / or screening of at least tuberculosis caused by the presence of the pathogen Mycobacterium tuberculosis, at least resistant to rifampicin.

[0014] Further, the Mycobacterium tuberculosis infection includes: primary infection, secondary infection, extrapulmonary infection.

[0015] In a third aspect, the present application provides a detection kit for differential diagnosis and / or screening of rifampicin-resistant pulmonary tuberculosis, the kit comprising reagents for detecting metabolic markers, instructions, etc., wherein the metabolic marker is the rifampicin-resistant pulmonary metabolic marker of the first aspect.

[0016] In a fourth aspect, the present application provides a use of one or more of the metabolic markers, Dihydroberberine, Dimethyl Sulfoxide, Choline, N,N-Dimethylarginine, Pyridoxine, D-Alanyl-D-alanine, Proline-Hydroxyproline, 4-(Cytisin-12-Amido)-Benzoic Acid, Phosphatidylethanolamine Lyso 20, (2-Oxo-2,3-Dihydro-1H-Indol-3-Yl) Acetic Acid, Cycloserine, 4-Pyridoxic Acid, Difluoro(Perfluoromethoxy) Acetic Acid, Beta-Alanine, in the manufacture of a medicament for treating rifampicin-resistant tuberculosis.

[0017] In a fifth aspect, the present application provides a use of one or more of the metabolic markers, 1-Palmitoyl-2-Oleoyl-3-Linoleoyl-Rac-Glycerol, 2-Acetoxy-4-Pentadecylbenzoic Acid, 1-Cinnamoylpyrrolidine, C18-Sphingosine, Erucamide, Beta-Hydroxymyristic Acid, in the manufacture of a medicament for treating rifampicin-resistant tuberculosis.

[0018] In a sixth aspect, the present application provides a screening method for identifying and / or screening rifampicin-resistant tuberculosis, the method comprising the following steps:

[0019] S1. Urine collection and preservation: collecting urine sample from the middle section of the morning urine of the subject;

[0020] S2. Extraction of metabolites in urine: the sample is subjected to freeze concentration drying, precipitation, centrifugation, and the supernatant is vacuum dried and then subjected to chromatography and mass spectrometry analysis to obtain peak alignment, retention time correction and extraction peak area;

[0021] S3. Comparison of the data obtained by the above chromatography and mass spectrometry with the public databases HMDB, MassBank and standard library baseDeepBP, and normalization of the data;

[0022] S4. Multivariate statistical analysis: PCA, PLS-DA and OPLS-DA methods are used to statistically analyze the normalized data and the data between the healthy control group, and to screen the urine metabolic markers of rifampicin-resistant pulmonary tuberculosis patients.

[0023] S5. Identification of differential metabolic markers: evaluated by area under the curve (AUC) and confusion matrix of metabolic markers.

[0024] Further, the stability evaluation of the mass spectrometry method is evaluated by quality control samples (QC).

[0025] Further, the quality control sample is a sample in which the signal peak with a relative standard deviation (RSD) of ≥30% is removed.

[0026] Further, the relative deviation refers to the relative standard deviation (RSD) after superimposed comparison of the total ion current chromatogram in the positive ion mode and the negative ion mode.

[0027] Further, the data normalization refers to obtaining information including compound retention time, mass-to-charge ratio and peak intensity by using accurate mass number matching (mass tolerance <20ppm) and secondary spectrum matching (mass tolerance <0.02Da) and database comparison.

[0028] Further, the data normalization also includes missing value processing.

[0029] Further, the missing value processing refers to deleting ion peaks with more than 50% missing values in the sample within the group.

[0030] Further, the screening criteria are VIP value >1.0, |log2FoldChange|≥0.26, and P<0.05.

[0031] In a seventh aspect, the present application provides a detection system for differential diagnosis and / or screening of rifampicin-resistant tuberculosis, the system comprising a data input module, a data processing module and a result output module.

[0032] Further, the data input is Dihydroberberine, Dimethyl Sulfoxide, 1-Palmitoyl-2-Oleoyl-3-Linoleoyl-Rac-Glycerol, Choline, N,N-Dimethylarginine, Pyridoxine, D-Alanyl-D-alanine, Proline-Hydroxyproline, 2-Acetoxy-4-Pentadecylbenzoic Acid, 1-Cinnamoylpyrrolidine, 4-(Cytisin-12-Amido)-Benzoic Acid, Phosphatidylethanolamine Lyso 20, C18-Sphingosine, Erucamide, Beta-Hydroxymyristic Acid, (2-Oxo-2,3-Dihydro-1H-Indol-3-Yl)Acetic Acid, Cycloserine, 4-Pyridoxic Acid, Difluoro(Perfluoromethoxy)Acetic Acid, Beta-Alanine of the urine metabolite extract of the subject obtained by chromatography and mass spectrometry as input data.

[0033] Further, the data processing is the detection of the abundance value and the detection of the log2FoldChange value of the abundance value of the input data.

[0034] Further, the data output is the log2FoldChange value of the abundance of the input data; that is, when the abundance of one or more of the metabolic markers Dihydroberberine, Dimethyl Sulfoxide, Choline, N,N-Dimethylarginine, Pyridoxine, D-Alanyl-D-alanine, Proline-Hydroxyproline, 4-(Cytisin-12-Amido)-Benzoic Acid, Phosphatidylethanolamine Lyso 20, (2-Oxo-2,3-Dihydro-1H-Indol-3-Yl) Acetic Acid, Cycloserine, 4-Pyridoxic Acid, Difluoro(Perfluoromethoxy) Acetic Acid, Beta-Alanine is < -0.26, that is, the abundance is down-regulated, indicating that the patient has a high risk of rifampicin-resistant pulmonary tuberculosis; when the abundance of one or more of the metabolic markers 1-Palmitoyl-2-Oleoyl-3-Linoleoyl-Rac-Glycerol, 2-Acetoxy-4-Pentadecylbenzoic Acid, 1-Cinnamoylpyrrolidine, C18-Sphingosine, Erucamide, Beta-Hydroxymyristic Acid is > 0.26, that is, the abundance is up-regulated, indicating that the patient has a high risk of rifampicin-resistant pulmonary tuberculosis. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1 OPLS-DA permutation plot of the cation mode for the rifampicin-resistant pulmonary tuberculosis patient group and the sensitive pulmonary tuberculosis patient group;

[0036] Figure 2Score plot of OPLS-DA model in cation mode for the group of patients with rifampicin-resistant pulmonary tuberculosis and the group of patients with sensitive pulmonary tuberculosis.

[0037] Figure 3 Score plot of PCA model in cation mode for the group of patients with rifampicin-resistant pulmonary tuberculosis and the group of patients with sensitive pulmonary tuberculosis.

[0038] Figure 4 Permutation plot of PLS-DA model in cation mode for the group of patients with rifampicin-resistant pulmonary tuberculosis and the group of patients with sensitive pulmonary tuberculosis.

[0039] Figure 5 Score plot of PLS-DA model in cation mode for the group of patients with rifampicin-resistant pulmonary tuberculosis and the group of patients with sensitive pulmonary tuberculosis.

[0040] Figure 6 Permutation plot of OPLS-DA model in anion mode for the group of patients with rifampicin-resistant pulmonary tuberculosis and the group of patients with sensitive pulmonary tuberculosis.

[0041] Figure 7 Score plot of OPLS-DA model in anion mode for the group of patients with rifampicin-resistant pulmonary tuberculosis and the group of patients with sensitive pulmonary tuberculosis.

[0042] Figure 8 Score plot of PCA model in anion mode for the group of patients with rifampicin-resistant pulmonary tuberculosis and the group of patients with sensitive pulmonary tuberculosis.

[0043] Figure 9 Permutation plot of PLS-DA model in anion mode for the group of patients with rifampicin-resistant pulmonary tuberculosis and the group of patients with sensitive pulmonary tuberculosis.

[0044] Figure 10 Score plot of PLS-DA model in anion mode for the group of patients with rifampicin-resistant pulmonary tuberculosis and the group of patients with sensitive pulmonary tuberculosis.

[0045] Figure 11 Volcano plot of urine metabolites in negative ion mode for the group of patients with rifampicin-resistant pulmonary tuberculosis and the group of patients with sensitive pulmonary tuberculosis.

[0046] Figure 12 Volcano plot of urine metabolites in positive ion mode for the group of patients with rifampicin-resistant pulmonary tuberculosis and the group of patients with sensitive pulmonary tuberculosis.

[0047] Figure 13 Heat map of urine metabolites for the group of patients with rifampicin-resistant pulmonary tuberculosis and the group of patients with sensitive pulmonary tuberculosis.

[0048] Figure 14 ROC curve. DETAILED DESCRIPTION

[0049] The rifampicin-resistant tuberculosis of the present application is that the Mycobacterium tuberculosis infected by the tuberculosis patient has at least the drug resistance to rifampicin.

[0050] The specific embodiments of the present application are further described below. It should be noted that the description of the embodiments is for the purpose of helping understand the present application, but does not constitute a limitation on the present application. In addition, the technical features involved in the following described embodiments can be combined with each other as long as they do not conflict with each other.

[0051] The experimental methods in the following examples are all conventional methods, and the experimental materials used in the following examples are all commercially available unless otherwise specified.

[0052] The term "diagnosis of rifampicin-resistant tuberculosis" in the present application is performed according to the new definition of WHO 2021, including: ① rifampicin-resistant tuberculosis (RR-TB): refers to tuberculosis caused by Mycobacterium tuberculosis infected in patients, which is confirmed by in vitro drug sensitivity test to be resistant to single rifampicin or rifampicin and drugs other than isoniazid; ② multidrug-resistant tuberculosis (MDR-TB): refers to tuberculosis caused by Mycobacterium tuberculosis infected in patients, which is confirmed by in vitro drug sensitivity test to be resistant to isoniazid and rifampicin at least at the same time; ③ pre-extensively drug-resistant tuberculosis (pre-XDR-TB): tuberculosis caused by MTB strains resistant to any fluoroquinolone drugs (including levofloxacin, moxifloxacin) that meet the definition of MDR / RR-TB; ④ extensive drug-resistant tuberculosis (XDR-TB): refers to tuberculosis caused by Mycobacterium tuberculosis infected in patients, which is confirmed by in vitro drug sensitivity test to be resistant to any quinolone drugs in addition to at least isoniazid and rifampicin, and bedaquiline or linezolid in group A drugs. The rifampicin-resistant pulmonary tuberculosis patients included in the present application all meet the above diagnostic criteria; and the rifampicin-resistant patients described in the present application all involve the types of rifampicin-resistant tuberculosis (RR-TB), multidrug-resistant tuberculosis (MDR-TB), pre-extensively drug-resistant tuberculosis (pre-XDR-TB) and extensive drug-resistant tuberculosis (XDR-TB).

[0053] Example 1 Screening and identification of urine metabolites

[0054] 1. Inclusion and grouping of research subjects

[0055] A total of 80 people aged 18-65 years old, including sensitive pulmonary tuberculosis and rifampicin-resistant pulmonary tuberculosis patients admitted to Beijing Chest Hospital, Capital Medical University, all of whom signed informed consent forms and recorded their clinical case information in detail. The specific grouping is as follows:

[0056] (1) Sensitive tuberculosis group (40 cases): The diagnosis of pulmonary tuberculosis refers to the health industry standard of the People's Republic of China (WS288-2017). All enrolled patients with pulmonary tuberculosis meet the diagnostic criteria of pulmonary tuberculosis and no drug resistance is found in drug resistance detection. Anyone who meets the following items is diagnosed with pulmonary tuberculosis: ① Two sputum smears positive for acid-fast bacilli; ② 1 sputum smear positive for acid-fast bacilli, chest imaging with typical active tuberculosis lesions; ③ 1 sputum smear positive for acid-fast bacilli, and 1 sputum smear positive for mycobacterium culture, and the strain identification is Mycobacterium tuberculosis complex; ④ Chest imaging with typical active tuberculosis lesions, sputum smear negative, but mycobacterium culture positive and strain identification is Mycobacterium tuberculosis complex; ⑤ Chest imaging with typical active tuberculosis lesions, Mycobacterium tuberculosis nucleic acid detection positive; ⑥ Lung histopathology and positive etiology. Exclusion criteria: drug resistance found during subsequent treatment.

[0057] (2) Rifampicin-resistant tuberculosis group (40 cases): ① Phenotype or molecular drug susceptibility test (DST) confirmed rifampicin-resistant tuberculosis patients; ② According to the treatment principles of WHO and China's guidelines and consensus on drug-resistant tuberculosis; ③ No severe heart, liver, kidney, and lung dysfunction.

[0058] 2. Urine collection and preservation

[0059] Collect 50 mL of midstream urine from the morning urine of the subject, place it in a 50 mL centrifuge tube, centrifuge at 1500 g for 10 min at 4°C, transfer the supernatant to a new centrifuge tube, discard the precipitate, and store at -80°C until use. All samples are renamed in the form of groups and serial numbers to protect the privacy of the subjects.

[0060] 3. Metabolomics technology screening of differential metabolites

[0061] 3.1 Metabolite extraction

[0062] Thaw the sample at 4°C, take 1 mL of each sample for freeze-concentration drying, add 1 mL of pre-cooled methanol / acetonitrile / water (2:2:1, v / v / v) respectively, ultrasonic in ice bath for 60 min, incubate at -20°C for 1 h to precipitate protein, centrifuge at 14000 g for 20 min at 4°C, take the supernatant for vacuum drying. Add 100 μL of acetonitrile-water solution (1:1, v / v) for reconstitution before mass spectrometry detection, centrifuge at 14000 g for 15 min at 4°C, take the supernatant for sample analysis. Take 5 μL of all processed samples for mixing to prepare a quality control sample (QC) for evaluating the stability of the experimental method.

[0063] 3.2 Chromatographic separation

[0064] Samples were separated using a SHIMADZU Nexera X2 LC-30AD ultra-high pressure liquid chromatograph (Shimadzu, Japan). The chromatographic column used was an ACQUITYUPLC HSS T3 (2.1 × 150 mm, 1.8 μm) (Waters, Milford, MA, USA). The mobile phases were: Solution A consisted of water containing 25 mM ammonium acetate and 25 mM ammonia; Solution B consisted of 100% acetonitrile. Samples were placed in an autosampler at 4 °C, with a column temperature of 25 °C, a flow rate of 0.3 mL / min, and an injection volume of 5 μL. The liquid chromatography gradient was as follows: 0-0.5 min, solution B was maintained at 95%; 0.5-7 min, solution B linearly changed from 95% to 65%; 7-9 min, solution B linearly changed from 65% to 40%; 9-10 min, solution B was maintained at 40%; 10-11.1 min, solution B linearly changed from 40% to 95%; 11.1-16 min, solution B was maintained at 95%. One QC sample was set up every 8 experimental samples in the sample cohort to detect and evaluate the stability and repeatability of the system.

[0065] 3.3 Mass Spectrometry Sets

[0066] Ultra-high performance liquid chromatography-quadrupole orbital trap mass spectrometry (UPLC-Q-Exactive-Orbitrap-MS / MS) was used.

[0067] (Thermo Scientific, USA) Mass spectrometry was performed in both positive and negative ion modes. ESI source parameters are as follows:

[0068] Source Temperature 600℃, Ion Source Gas1 (GAS1): 60, Ion Source Gas2 (GAS2): 60, Curtain Gas (CUR): 30, Ion Spray Voltage Floating (ISVF) ±5500V. TOF MS scan m / z range: 60-1200 Da, product ion scan m / z range: 25-1200 Da, TOF MS scan accumulation time 0.15 s / spectra, product ion scan accumulation time 0.03 s / spectra; secondary mass spectrum was obtained by Information Dependent Acquisition (IDA), and high sensitivity mode was used, Declustering potential (DP): ±60V, Collision Energy: 30eV, IDA settings as follows: Exclude isotopes within 4 Da, Candidate ions to monitor per cycle: 6.

[0069] 3.4 Quality control analysis

[0070] The QC sample mass spectrum total ion chromatogram in positive ion mode and negative ion mode was compared by spectrum superposition, and the relative standard deviation (RSD) of each signal peak of the QC sample was calculated. The signal peaks with RSD exceeding 30% were removed.

[0071] 3.5 Data preprocessing

[0072] The original data was processed by MSDIAL software for peak alignment, retention time correction and peak area extraction. Metabolite structure identification was performed by accurate mass number matching (mass tolerance <20ppm) and secondary spectrum matching (mass tolerance <0.02 Da), and public databases such as HMDB, MassBank and self-built standard library baseDeepBP were searched to obtain information including compound retention time, mass-to-charge ratio and peak intensity. Missing value processing was performed by deleting ion peaks with more than 50% missing values in the group. The positive and negative ion data were normalized by total peak area, and the positive and negative ion peaks were integrated and subjected to pattern recognition by R software. After UV pretreatment, the data was subjected to subsequent data analysis.

[0073] 4. Multivariate statistical analysis

[0074] The obtained data were subjected to PCA, PLS-DA and OPLS-DA analysis. In combination with univariate analysis, including Fold Change Analysis (FC Analysis), T-test, and Volcano Plot, etc., the screening conditions were set as |log2FoldChange|≥0.26 and P<0.05, so as to screen potential differential metabolic markers. The analysis results are shown in Table 1. Figures 1 to 12

[0075] 5. Differential metabolite identification

[0076] Based on the determination of a total of 80 cases, including 40 rifampicin-resistant tuberculosis patients and 40 sensitive tuberculosis patients, the urine metabolite results were analyzed, and the metabolic markers were identified (as shown in Table 1).

[0077] Table 1. List of differential metabolites

[0078]

[0079]

[0080] The SPSS software was used for receiver operating curve (ROC) calculation based on the logistic model, and finally 20 differential metabolites were identified. The area under the curve (AUC) of the 20 urine metabolic markers was obtained, and in the 95% confidence interval, the AUC of the 20 urine metabolic markers was between 0.703125-0.93875 (as shown in Table 2).

[0081] ​By comparing the metabolite abundance of rifampicin-resistant tuberculosis individuals and sensitive tuberculosis individuals, it is concluded that the down-regulation of Dihydroberberine, Dimethyl Sulfoxide, Choline, N,N-Dimethylarginine, Pyridoxine, D-Alanyl-D-alanine, Proline-Hydroxyproline, 4-(Cytisin-12-Amido)-Benzoic Acid, Phosphatidylethanolamine Lyso 20, (2-Oxo-2,3-Dihydro-1H-Indol-3-Yl) Acetic Acid, Cycloserine, 4-Pyridoxic Acid, Difluoro(Perfluoromethoxy) Acetic Acid, and Beta-Alanine metabolites indicates a high risk of rifampicin-resistant tuberculosis.

[0082] 1-Palmitoyl-2-Oleoyl-3-Linoleoyl-Rac-Glycerol, 2-Acetoxy-4-Pentadecylbenzoic Acid, 1-Cinnamoylpyrrolidine, C18-Sphingosine, Erucamide, and Beta-Hydroxymyristic Acid metabolite abundance up-regulation indicates a high risk of rifampicin-resistant tuberculosis (as shown in Table 2).

[0083] Table 2 Metabolic marker data

[0084]

[0085] The expression levels of 20 urine differential metabolites and 80 samples including 40 rifampicin-resistant tuberculosis (R) and 40 sensitive tuberculosis (S) were subjected to cluster analysis. The average value of the same sample metabolite expression was used as the benchmark. If the expression value was higher than the average value, it was positive and marked as red. Otherwise, if the expression value was lower than the average value, it was negative and marked as blue. The color depth represented the difference between the metabolite expression and the average value (see Figure 13 ).

[0086] Example 2 Construction of a rifampicin-resistant tuberculosis diagnosis model

[0087] Through sample collection and data sorting, a prediction model was constructed for 20 urine differential metabolic markers for testing. Through confusion matrix evaluation, a total of 80 samples (40 rifampicin-resistant tuberculosis patients and 40 sensitive tuberculosis patients) were used. The model prediction true positive (TP) value was 38, the false positive (FP) value was 2, the true negative (TN) value was 37, and the false negative (FN) value was 3.

[0088] The accuracy of the model was 93.75%; the precision was 95%; the sensitivity was 92.68%; and the specificity was 94.87%. The F1-Score value of the model was 0.9383, and the model output was good (as shown in Table 3).

[0089] Table 3 Confusion matrix

[0090] N＝80 Predicted negative Predicted positive True negative 37 2 True positive 3 38

[0091] The partial least squares algorithm was used to model and draw the ROC curve according to the 20 urine differential metabolic markers screened. The classification performance of the rifampicin-resistant tuberculosis detection model obtained is shown in Figure 14 ). According to the order of the urine differential metabolic markers in Table 1, the AUC values of the first 2, 3, 5, 7, 10, and 20 urine differential metabolic marker combinations were 0.849, 0.862, 0.89, 0.907, 0.922, and 0.931, respectively. It can be seen that the 20 urine metabolic markers screened have high accuracy for rifampicin-resistant tuberculosis detection and are of great significance for the diagnosis of rifampicin-resistant tuberculosis.

Claims

1. A combination of metabolic markers for differentiating Rifampicin resistant tuberculosis and sensitive pulmonary tuberculosis, the combination of metabolic markers being Dihydroberberine, Dimethyl Sulfoxide, 1 -Palmitoyl-2-Oleoyl-3-Linoleoyl-Rac-Glycerol, Choline, N,N-Dimethylarginine, Pyridoxine, D-Alanyl-D-alanine, Proline-Hydroxyproline, 2-Acetoxy-4-Pentadecylbenzoic Acid, 1 -Cinnamoylpyrrolidine, 4-(Cytisin-12-Amido)-Benzoic Acid, Phosphatidylethanolamine Lyso 20, C18-Sphingosine, Erucamide, Beta-Hydroxymyristic Acid, (2-Oxo-2,3-Dihydro-1 H-Indol-3-Yl)Acetic Acid, Cycloserine, 4-Pyridoxic Acid, Difluoro(Perfluoromethoxy)Acetic Acid and Beta-Alanine.

2. Use of a combination of metabolic markers as claimed in claim 1 for the preparation of a medicament for differentiating Rifampicin resistant tuberculosis and sensitive pulmonary tuberculosis.

Images