Rheumatism syndrome biomarker, diagnosis / prediction model and construction method of diagnosis / prediction model

Abstract

The invention relates to the technical field of prediction models, and particularly discloses a pain and rheumatism syndrome biomarker, a diagnosis / prediction model and a construction method thereof. The pain and rheumatism syndrome biomarker comprises at least one of acetyl vanillin, cyclic adenosine monophosphate, methyl vanillate and uridine succinic acid; the pain and rheumatism syndrome biomarker comprises a pain and rheumatism syndrome diagnosis marker and a pain and rheumatism syndrome recurrence prediction marker; the pain and rheumatism syndrome diagnosis marker comprises at least one of acetyl vanillin, cyclic adenosine monophosphate and methyl vanillate; the pain and rheumatism syndrome recurrence prediction marker comprises cyclic adenosine monophosphate and uridine succinic acid. According to the application, a diagnosis model of the pain and rheumatism syndrome can be constructed on the basis of the expression levels of the three plasma metabolites, namely the acetylvanillin, the cyclic adenosine monophosphate and the methyl vanillate; based on the plasma expression level of cyclic adenosine monophosphate and uridine succinic acid, a gout recurrence prediction model can be developed, and high prediction accuracy is achieved.

Description

Technical Field

[0001] The present application relates to the technical field of prediction models, and in particular to a gouty rheumatic disease biomarker, a diagnostic / predictive model, and a method for constructing the same. Background Art

[0002] Gout (GA) is an inflammatory disease characterized by hyperuricemia and the accumulation of monosodium urate (MSU) in joints and soft tissues. However, traditional treatments have limitations in effectively controlling gout comorbidities, such as metabolic syndrome, which often further complicate the clinical course of gout. According to statistics, the gout recurrence rate for patients using the globally recommended targeted treatment, urate-lowering treatments (ULTs), to maintain serum uric acid levels (SUA) ≤6.0 mg / dL, is 33.3% per patient per year. For patients not receiving ULTs, the gout recurrence rate is 49.1% per patient per year. Furthermore, only one-third to one-half of gout patients receive ULTs, and less than one-half of these patients receive treatment according to their doctor's instructions. This suggests that effective management of gout recurrences remains a significant challenge.

[0003] According to Traditional Chinese Medicine (TCM) theory, dampness is considered the primary cause of metabolic abnormalities and recurrent disease. The clinical manifestations of dampness are known as dampness syndrome (DS). However, research on dampness syndrome, particularly in patients with gouty arthritis and DS, remains limited. For example, the characteristic markers of dampness syndrome and the mechanisms by which dampness syndrome leads to gout attacks or relapses remain unclear. Therefore, identifying reliable biomarkers and establishing diagnostic and predictive models are crucial for dampness-syndrome gout. Summary of the Invention

[0004] The purpose of this application is to overcome the deficiencies of the above-mentioned prior art and to provide a gouty rheumatic disease biomarker, a diagnostic / predictive model and a method for constructing the same.

[0005] To achieve the above objectives, the technical solutions adopted in this application are:

[0006] The present application provides a gouty rheumatism biomarker, which includes at least one of acetylvanillone, cyclic adenosine monophosphate, methyl vanillate and uridine succinate.

[0007] As a preferred embodiment of the gouty rheumatic syndrome biomarker described in the present application, the gouty rheumatic syndrome biomarker includes a gouty rheumatic syndrome diagnostic marker and a gouty rheumatic syndrome recurrence prediction marker;

[0008] The gouty rheumatic syndrome diagnostic markers include at least two of acetylvanillone, cyclic adenosine monophosphate, and methyl vanillate;

[0009] The predictive markers for relapse of gouty rheumatic disease include cyclic adenosine monophosphate and uridine succinate.

[0010] After extensive research and experiments, the inventors of the present application discovered that a diagnostic model for gouty rheumatism can be constructed based on the expression levels of three plasma metabolites: acetovanillone, cyclic adenosine monophosphate (Cyclic AMP), and methyl-vanillate.

[0011] Among them, the ACU of acetovanillone was 0.915, the ACU of methyl-vanillate was 0.826, and the ACU of cyclic adenosine monophosphate (Cyclic-AMP) was 0.794.

[0012] A gout recurrence prediction model can be developed based on the plasma expression levels of cyclic adenosine monophosphate (CAMP) and uridine succinic acid (URS).

[0013] As a preferred embodiment of the gouty rheumatic syndrome biomarker described in the present application, the gouty rheumatic syndrome diagnostic markers are acetylvanillone and cyclic adenosine monophosphate.

[0014] This application uses acetylvanillone and cyclic adenosine monophosphate as diagnostic markers for rheumatic gout, with an average OOB value of 0.158 in the training data set and the highest average accuracy of 84.2% in the test data set.

[0015] As a preferred embodiment of the gouty rheumatic syndrome biomarker described in the present application, the gouty rheumatic syndrome recurrence prediction marker also includes clinical indicators.

[0016] As a preferred embodiment of the gouty rheumatism biomarker described in the present application, the clinical indicators include creatine kinase isoenzyme or low-density lipoprotein cholesterol.

[0017] Gouty-rheumatic relapse prediction markers include all possible combinations of four indicators: cyclic adenosine monophosphate, uridine succinate, creatine kinase isoenzymes, and low-density lipoprotein cholesterol, totaling 15 combinations. The model constructed using cyclic adenosine monophosphate and creatine kinase isoenzymes was the best, demonstrating an accuracy of 67.39% and an AUC of 0.803 in the validation cohort (cohort 2). Using these gouty-rheumatic relapse prediction markers, a gouty-rheumatic relapse prediction model can be constructed.

[0018] The present application also provides the use of the above-mentioned gouty rheumatism biomarkers in gouty rheumatism diagnosis products and / or gouty rheumatism recurrence prediction products.

[0019] The present application also provides a kit for diagnosing or predicting recurrence of gouty rheumatism, which includes a reagent for detecting the gouty rheumatism biomarker.

[0020] Preferably, the reagents for detecting the gouty rheumatic syndrome biomarkers include reagents for detecting gouty rheumatic syndrome diagnostic markers and reagents for detecting gouty rheumatic syndrome relapse prediction markers.

[0021] The present application also provides a method for constructing a gouty rheumatism diagnosis model, comprising the following steps:

[0022] S1. Metabolite differences among healthy subjects, patients with rheumatic gout, and patients with non-rheumatic gout were analyzed, and the AUC values ​​of the metabolites were calculated to screen out acetylvanoic acid, cyclic adenosine monophosphate, and methyl vanillate.

[0023] S2. Use the random forest algorithm to read data from healthy individuals, patients with rheumatic gout, and patients with non-rheumatic gout. Divide the data into a training set and a validation set. Use at least two of acetylvanoyl, cyclic adenosine monophosphate, and methyl vanillate as variables to set model parameters. After running the model, extract the out-of-bag error of the training set and the accuracy of the validation set.

[0024] The combination of acetylvanoyl, cyclic adenosine monophosphate, and methyl vanillate was used as the variable, and the parameters of the model were set as follows: ntree = 2000, mtry = 2, proximity = TRUE, importance = TRUE, keep.inbag = TRUE;

[0025] The combination of acetylvanillone and cyclic adenosine monophosphate, or the combination of acetylvanillone and methyl vanillate, or the combination of cyclic adenosine monophosphate and methyl vanillate is used as a variable, and the parameters of the model are set as follows: ntree=2000, mtry=1, proximity=TRUE, importance=TRUE, keep.inbag=TRUE.

[0026] The present application also provides a method for constructing a recurrence prediction model for gout-rheumatic syndrome, comprising the following steps:

[0027] S1. OPLS-DA algorithm was used to screen the differential metabolites between recurrent gouty rheumatic syndrome and non-recurrent gouty rheumatic syndrome according to the thresholds of VIP>1.5 and P<0.05 to obtain cyclic adenosine monophosphate and uridine succinate;

[0028] S2. Different combinations of cyclic adenosine monophosphate, uridine succinate, and differential clinical indicators were used to build models using the logistic regression method. The models were then validated and the accuracy and AUC values ​​were calculated to evaluate the predictive performance of different combinations.

[0029] Compared with the prior art, this application has the following beneficial effects:

[0030] The present application provides a biomarker for gouty rheumatism, a diagnostic / predictive model, and a method for constructing the same. The present application can construct a diagnostic model for gouty rheumatism based on the expression levels of three plasma metabolites: acetovanillone, cyclic adenosine monophosphate (Cyclic AMP), and methyl vanillate; and can develop a gout recurrence prediction model with high accuracy based on the plasma expression levels of cyclic adenosine monophosphate (Cyclic AMP) and creatine kinase isoenzyme (CKMB). BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Figure 2 is the plasma OPLS-DA analysis and differential metabolite analysis;

[0032] Figure 2 This is the analysis diagram of common differential metabolites;

[0033] Figure 3 This is the analysis chart of plasma metabolites specific to gouty rheumatic syndrome (GA-DS);

[0034] Figure 4 Figure 2 shows the metabolic differences and predictive modeling results of relapse in gouty rheumatic syndrome (GA-DS). DETAILED DESCRIPTION

[0035] In order to better illustrate the purpose, technical solutions and advantages of this application, this application will be further described below with reference to the accompanying drawings and specific embodiments.

[0036] In the following examples and comparative examples, the experimental methods used are conventional methods unless otherwise specified, and the materials, reagents, etc. used are all commercially available unless otherwise specified, and the components and raw materials used in each parallel experiment are all of the same kind.

[0037] Example 1: Screening of predictive markers for relapse of gouty rheumatic syndrome

[0038] 1) Experimental design:

[0039] This study adopted a prospective research method. Based on the gouty arthritis rheumatism research cohort constructed from clinical research, age-matched gouty rheumatism and non-rheumatism populations were selected for comparative analysis of plasma metabolites with those of healthy people (cohort 1). At the same time, based on the recurrence within 24 weeks, the plasma metabolites of age-matched gouty rheumatism recurrence and non-recurrence populations were selected for further analysis and verification (cohort 2). This study plan was approved by the Ethics Committee of Guangdong Provincial Hospital of Traditional Chinese Medicine, with approval numbers (BF2020-193-01 and BF2021-235-01).

[0040] Among them, the clinical characteristics of the participants in cohort 1 were summarized, which included three groups: healthy control group (HC, n = 24), gouty arthritis dampness syndrome group (DS, n = 30) and non-dampness syndrome group (NDS, n = 30).

[0041] Cohort 2: Gouty rheumatic syndrome relapse group (GA-R, n=22) and non-relapse group (GA-NR, n=24).

[0042] 2) Research content:

[0043] The following information was collected from the patients: (1) general information: including name, gender, age, telephone number, medical card number, height, weight, body mass index (BMI), etc.; (2) relapse information: including relapse date and relapse frequency; (3) TCM dampness syndrome scale: a collection form for information on core symptoms of TCM dampness syndrome. Questions that the patients could not understand during the scale filling were guided by rheumatologists, and tongue and pulse were determined by rheumatologists; (4) laboratory indicators: including erythrocyte sedimentation rate, C-reactive protein, three renal function tests, two liver function tests, three cardiac enzyme tests, fasting blood sugar, blood routine test, and urine routine test.

[0044] 3) Sample collection and processing:

[0045] Plasma was collected from patients with gouty arthritis and healthy donors for clinical research. Blood samples from study participants were collected using EDTA-K2 anticoagulant tubes. Samples were immediately transported to the laboratory, centrifuged at 3000 rpm for 10 minutes, and the upper plasma layer was separated. The plasma was transferred to cryovials and stored at -80°C. All participants in this study provided written informed consent. All samples were stored at the Biological Resource Center of Guangdong Provincial Hospital of Traditional Chinese Medicine according to standard operating procedures.

[0046] 4) Plasma metabolite extraction:

[0047] Using BGI's HM700 targeted metabolomics assay, 20 μL of plasma and 20 μL of standard were mixed with 120 μL of sample release reagent, shaken at 1200 rpm for 30 minutes, and centrifuged at 18,000 g for 30 minutes at 4°C. 30 μL of supernatant was then transferred to a 96-well plate. Subsequently, 20 μL of derivatization reagent and 20 μL of EDC working solution were added. The plate was sealed with aluminum film and incubated at 1200 rpm and 40°C for 60 minutes. The incubated mixture was centrifuged at 4000 g for 5 minutes at 4°C. Then, 30 μL of supernatant was transferred to a new 96-well plate, and 90 μL of sample diluent was added to each well. After shaking at 600 rpm for 10 minutes, the plate was centrifuged at 4000 g for 30 minutes at 4°C. The plate was sealed and prepared for analysis.

[0048] 5) UPLC-MS analysis (ultra-performance liquid chromatography-mass spectrometry analysis):

[0049] A Waters ACQUITY UPLC I-Class Plus system (Waters, USA) combined with a QTRAP6500Plus high-sensitivity mass spectrometer (SCIEX, USA) was used for metabolite separation and quantification.

[0050] Chromatographic conditions: Chromatographic separation was performed on a BEH C18 column (2.1 mm×10 cm, 1.7 μm, Waters), with the mobile phase consisting of 0.1% formic acid in water (solvent A) and 30% isopropanol in acetonitrile (solvent B).

[0051] The gradient elution program is as follows:

[0052] 0.00-1.00 min: 5% B; 1.00-5.00 min: 5%-30% B; 5.00-9.00 min: 30%-50% B; 9.00-11.00 min: 50%-78% B; 11.00-13.50 min: 78%-95% B; 13.50-14.00 min: 95%-100% B (flow rate: 0.400 ml / min); 14.00-16.00 min: 100% B (flow rate: 0.600 ml / min); 16.00-18.00 min: 5% B (flow rate: 0.400 ml / min).

[0053] Mass spectrometry conditions: A QTRAP 6500 Plus mass spectrometer equipped with an ESI Turbo ion spray interface was used, and the ion source parameters were set as follows: ion source temperature: 400°C; ion spray voltage (IS): 4500 V in positive mode and -4500 V in negative mode; ion source gases I (GS1), II (GS2), and curtain gas (CUR) were set to 60, 60, and 35 psi, respectively.

[0054] Multiple Reaction Monitoring (MRM) mode: MRM transitions are configured, including precursor and product ion pairs, collision energy (CE), declustering potential (DP), and retention time of target metabolites.

[0055] 6) Extraction and identification of metabolite ion peaks:

[0056] Metabolite ion peak extraction and identification were performed using Skyline software (version 21.1.0.146). This software generates a data matrix containing metabolite identification and quantification results. This data matrix information was then further analyzed and processed. Parameters set in Skyline included the mass of the monoisotopic peak, a mass tolerance of 0.6 Da, and a mass range of 50–1500 Da. Detailed information on the software's operation can be found on its official website: https: / / skyline.ms / project / home / software / Skyline / begin.view .

[0057] 7) Bioinformatics analysis of metabolomics data:

[0058] The results exported from Skyline were first normalized to obtain Z scores. Subsequent analyses were then performed using R Studio. To explore the differences in metabolomic profiles between samples in different groups, this study employed orthogonal partial least squares discriminant analysis (OPLS-DA) for cluster analysis or feature screening. After importing the raw metabolomics data and corresponding grouping information, the OPLS-DA model was implemented using the ropls R package. The number of predicted components (predI) and the number of orthogonal components (orthoI) were set to 1, Pareto normalization was used (scaleC="pareto"), and the number of cross-validation folds was used (crossvalI=nrow(dat)). A random seed (set.seed(1)) was set during model training to ensure reproducible results. To assess the stability and fit of the model, 1000 permutation tests (permI=1000) were performed, and permutation distribution plots were output. After model fitting, the VIP (Variable Importance in Projection) value of each variable was extracted to identify the metabolites that contributed most to the group differences. At the same time, an OPLS-DA score graph was drawn to show the distribution trend of the predicted component (t1) and the orthogonal component (to1) among different groups. The model fit goodness (R 2 X and R 2 Y) and predictive ability (Q 2 ) are also marked in the figure to comprehensively evaluate the model performance. In addition, by visualizing the R of the model in 1000 permutation tests 2 Y and Q 2 The distribution of the model is compared with the corresponding value of the actual model to evaluate whether the model is overfitting. Finally, the model score data, VIP results and permutation analysis results are combined.

[0059] In addition, logistic regression and random forest models were constructed using the glmnet and randomForest packages, respectively. Data visualization was achieved using the ggplot2 package in R. Box plots and volcano plots were also created using GraphPad Prism.

[0060] 8) Data entry and management:

[0061] A database was established using Excel, and data collected from patients during consultations and follow-up visits, as well as data extracted from the hospital information management system, were collated and entered. Information such as the patient's gender, age, consultation date, recurrence, TCM dampness syndrome score, and laboratory indicators were entered.

[0062] 9) Statistical analysis:

[0063] The measurement data that conform to the normal distribution are expressed as “mean ± standard deviation” Indicates that the measurement data that do not conform to the normal distribution are expressed as M(P 25 , P 75 ) indicates data from a randomized controlled trial. Statistical analyses were performed using GraphPad Prism v9 or the R statistical software package. Two-way ANOVA was used for multivariate comparisons between multiple groups; one-way ANOVA was used for multivariate comparisons between two groups. Univariate comparisons between two groups were performed using the independent sample t-test for normally distributed data and the nonparametric Mann-Whitney U test for data that did not. The Kruskal-Wallis H test was used for multi-group comparisons. Specific statistical methods used for each data set are detailed in the corresponding figure and table legends. Correlation analyses were performed using Spearman correlation analysis. P < 0.05 was considered statistically significant.

[0064] result:

[0065] (1) Plasma metabolic characteristics of patients with gouty rheumatic syndrome (GA-DS):

[0066] 1. Patients with gouty rheumatic syndrome have obvious abnormalities in metabolism and inflammation:

[0067] Table 1 summarizes the clinical characteristics of the participants in cohort 1, which included three groups: healthy control group (HC, n = 24), gouty arthritis dampness syndrome group (DS, n = 30), and non-dampness syndrome group (NDS, n = 30).

[0068] The body mass index (BMI) of the DS group was 25.59±2.86 kg / m 2 , which was significantly higher than that of the HC group (22.34±2.25kg / m 2 , P < 0.001) and NDS group (24.28 ± 2.99 kg / m 2 The serum uric acid (SUA) level in the DS group (460.93±125.43μmol / L) was significantly higher than that in the HC group (356.00±51.48μmol / L, P<0.01) and the NDS group (351.13±84.71μmol / L, P<0.01).

[0069] Key indicators of lipid metabolism showed that TC, TG, and LDL-C levels in the DS group were higher than those in the NDS group (5.10±0.95 mmol / L, P<0.01, 1.99 (1.49, 3.06) mmol / L, P<0.001, and 3.33 (2.55, 4.01) mmol / L, P<0.05, respectively). This suggests significant metabolic and physiological differences between the DS and NDS groups. Furthermore, the median erythrocyte sedimentation rate (ESR) in the DS group was significantly higher at 24.5 (15, 41.75) mm / h than in the NDS group (21.5 (14.75, 28.25) mm / h) (P<0.01). This suggests that DS has a potential impact on the metabolic and inflammatory status of patients with gouty arthritis. Further investigation is needed into the underlying mechanisms of these differences.

[0070] Table 1 Clinical characteristics

[0071]

[0072]

[0073] Note: Normality test results showed that age, BMI, SUA of each group, and eGFR and TC of DS and NDS groups were in accordance with normal distribution and homogeneity of variance (significance>0.05), expressed as (x±s). One-way ANOVA was used to compare the differences among the three groups. The other indicators did not conform to normal distribution (significance<0.05), expressed as M (P 25 , P 75 ) indicates that the differences among the three healthy groups were compared by Kruskal-Wallis H test, and the differences were statistically significant at P<0.05, *P<0.05, **P<0.01, ***P<0.001; the differences between the dampness syndrome group (DS) and the non-dampness syndrome group (NDS) were compared by Mann-Whitney U test, and the differences were statistically significant at P<0.05, #P<0.05, ##P<0.01, and ###P<0.001.

[0074] 2. There are significant differences in the plasma metabolic profiles among healthy people, patients with gouty rheumatic syndrome, and patients with gout without rheumatic syndrome:

[0075] To better understand the metabolic abnormalities associated with GA-DS, targeted metabolomics analysis was performed on plasma samples from all participants in cohort 1 (healthy controls (HC, n=24), gouty rheumatic syndrome (DS, n=30), and non-dampness syndrome (NDS, n=30)) using UPLC / MS. A total of 366 metabolites were detected and identified. Orthogonal partial least squares discriminant analysis (OPLS-DA) was then used for cluster analysis. The results showed that plasma metabolite expression profiles could effectively distinguish HC, GA-DS, and GA-NDS groups ( Figure 1 ).

[0076] like Figure 1 As shown in Figure A, the OPLS-DA model based on plasma metabolite expression data can clearly distinguish the GA-DS group from the HC group. The R 2 Y and Q 2 The Y values ​​were 0.934 and 0.801, respectively. The top three metabolites that contributed most to the model were uridine-5'-monophosphate, acetovanillone, and S-5'-adenosyl-L-homocysteine ​​(SAH).

[0077] In addition, if Figure 1 The volcano plot in Figure B is based on the contribution of metabolites to the model (VIP score) and their statistical differences between groups (FDR). Using a threshold of VIP score > 1 and -Log(FDR) > 2, 27 differential metabolites were identified, of which 16 were upregulated in the DS group ( Figure 1 The pink dots in B, such as acetovanillone and S-5'-adenosyl-L-homocysteine ​​(SAH), 11 down-regulated ( Figure 1 Similarly, there were significant differences in the plasma metabolite profiles between the GA-NDS group and the HC group ( Figure 1 Middle C, R 2 Y=0.923,Q 2 Y = 0.808), 51 differential metabolites were identified ( Figure 1 Moreover, although both the GA-DS and GA-NDS groups were gout patients, the two groups showed different plasma metabolite profiles ( Figure 1 Middle E, R 2 Y=0.876,Q 2 Y = 0.652), a total of 50 differential metabolites were identified (as shown in Table 2). Among them, 33 metabolites were increased in the GA-DS group, and 17 were decreased ( Figure 1 Middle F). These findings provide a comprehensive understanding of the metabolic characteristics of the GA-DS population.

[0078] Table 2 50 differential metabolites between DS and NDS

[0079]

[0080]

[0081]

[0082] 3. Compared with healthy people, patients with gouty rheumatic syndrome and gout without rheumatic syndrome have differences in 15 common metabolites:

[0083] In order to further analyze which metabolites are disease-specific (GA) and which are syndrome-specific (DS), this study Figure 1 Middle B, Figure 1 D and Figure 1 The differential metabolites identified in the volcano plot in Figure F were subjected to Venn diagram analysis.

[0084] The results are as follows Figure 2 As shown in Figure 5A, the results showed that there were 15 common differential metabolites between GA-DS and GA-NDS compared with HC, including 7 up-regulated and 8 down-regulated metabolites (see Figure 2 Among them, three metabolites, ortho-hydroxyphenylacetic acid, phthalic acid, and S-5'-adenosyl-L-homocysteine, were negatively correlated with eGFR, with Sperman correlation coefficients of -0.49, -0.53, and -0.59, respectively (see Figure 2 These metabolites were higher in gout patients, especially in the GA-DS group (see Figure 2 In addition, S-5'-adenosyl-L-homocysteine ​​was positively correlated with the inflammatory marker high-sensitivity C-reactive protein (hCRP) (r = 0.48) (see Figure 2 These shared disease-specific metabolites provide new insights into the pathogenesis of gout, and further studies are needed to elucidate their specific molecular roles in gout pathology.

[0085] Example 2: A diagnostic model for gout and rheumatism

[0086] After pairwise analysis of variance (nonparametric unpaired Mann-Whitney test), five metabolites, including acetovanillone, cyclic adenosine monophosphate (CAMP), methyl-vanillate, 5-aminoimidazole-4-carboxamide, and m-coumaric acid, were found to be elevated in GA-DS plasma. Figure 2 As shown in Figure A, there are five common metabolites between the GA-DS group and the HC and GA-NDS groups. These five metabolites are acetovanillone, cyclic adenosine monophosphate (CAMP), methyl vanillate, 5-aminoimidazole-4-carboxamide, and m-coumaric acid. Compared with the GA-NDS and HC groups, these metabolites were significantly increased in the GA-DS group, and the differences were statistically significant (P<0.05) ( Figure 3 Middle A).

[0087] Further correlation analysis with the patient's clinical indicators revealed that acetovanillone was positively correlated with SUA (r=0.50, P<0.05) ( Figure 3 In the middle B), cyclic adenosine monophosphate (CAMP) and m-coumaric acid (m-Coumaric acid) were negatively correlated with eGFR (r = -0.53, r = -0.54, P < 0.05), and positively correlated with hCRP (r = 0.48, P < 0.05) ( Figure 3 Middle C, Figure 3 (D), suggesting that these metabolites may be involved in inflammatory responses and may be associated with disease recurrence.

[0088] Currently, the clinical diagnosis of GA-DS mainly relies on scale scores and lacks objective diagnostic tools. In order to evaluate whether these five metabolites can be used to distinguish GA-DS from non-GA-DS groups (HC and GA-NDS in this study), their AUC values ​​were calculated. Then, the classification performance of different metabolites in the Corhort1 dataset (which only contains metabolomic data of GA-DS and GA-NDS) was evaluated using the multipleROC function in R language. The AUC (area under the curve) value was calculated to screen out the metabolites with the most predictive ability. The top three metabolites were acetovanillone (ACU = 0.915), methyl-vanillate (Methyl-vanillate) (ACU = 0.826) and cyclic adenosine monophosphate (Cyclic-AMP) (ACU = 0.794) ( Figure 3 Middle E).

[0089] During the model construction phase, the random forest algorithm using the R language randomForest package was used. After reading the data from Cohort 1, the seed number record was set and the datasets (GA-DS and GA-NDS) were split into training and validation sets in a 7:3 ratio. Four models were then constructed using the random forest algorithm, using the grouping information (GA-DS and GA-NDS) as the classification criterion: acetovanillone + methyl-vanillate + cyclic-AMP (A + M + C), acetovanillone + methyl-vanillate (A + M), acetovanillone + cyclic-AMP (A + C), and cyclic-AMP + methyl-vanillate (C + M).

[0090] Using A+C+M (Acetovanillone+Cyclic AMP+Methyl-vanillate) as the variables, set the model parameters to: ntree = 2000, mtry = 2, proximity = TRUE, importance = TRUE, keep.inbag = TRUE. Run the model 1000 times, and extract the out-of-bag error (OOB) of the training set and the accuracy of the test set after each run.

[0091] For the pairwise combination models, namely A+C, A+M, and C+M, the steps are the same as A+C+M, but the model parameters are ntree=2000, mtry=1, proximity=TRUE, importance=TRUE, and keep.inbag=TRUE.

[0092] After 10,000 modeling runs, it was found that the A+C model performed best, with the lowest OOB value and the highest accuracy. The average OOB value in the training dataset was 0.158 ( Figure 3 F), and achieved the highest average accuracy of 84.2% in the test dataset ( Figure 3 Middle G).

[0093] Example 3: A model for predicting recurrence of gout-rheumatic syndrome and its construction method

[0094] Traditional Chinese medicine theory believes that dampness syndrome is closely related to the recurrence of diseases such as gouty arthritis. This relationship was partially confirmed in the cohort of this study. Figure 4 As shown in Figure A, the relapse rates in the GA-DS group at 12 weeks and 24 weeks were 20.0% and 41.7%, respectively, which were higher than those in the GA-NDS group (13.6% and 27.8%), but these differences were not statistically significant (P>0.05).

[0095] To further understand the metabolic differences between patients with recurrent gout (GA-R, defined as at least one attack within 24 weeks) and those without recurrent gout (GA-NR, defined as no attack within 24 weeks), cohort 2 (including 24 GA-NR and 22 GA-R male patients, all of whom were in the gout intermittent period) was analyzed. The results showed that some clinical indicators of the two groups showed significant differences: the GA-R group had higher levels of SUA, TC, and LDL-C, while the CK-MB level was lower (see Figure 4 Middle B). These results suggest that the metabolic status of patients with gouty rheumatic syndrome is associated with the risk of relapse.

[0096] To identify potential predictors of gout recurrence, plasma samples from cohort 2 patients underwent the same targeted metabolomics analysis as cohort 1.

[0097] By combining OPLS-DA and statistical criteria (VIP>1.5, P<0.05), the differential metabolites between GA-R and GA-NR in the two cohorts were screened. Figure 4 As shown in the volcano plot in C, there are 27 differential metabolites in cohort 1 and 24 in cohort 2. The Venn diagram shows that there are only two metabolites that overlap between the two cohorts, namely cyclic adenosine monophosphate (cAMP) and uridine succinic acid (U), which are both increased in the GA-R group (see Figure 4 Middle D).

[0098] Subsequently, these two metabolites (cyclic adenosine monophosphate and uridine succinate) were combined with the patient's clinical indicators (creatine kinase isoenzyme and low-density lipoprotein cholesterol) into 15 different combinations. Using the data from cohort 1, 15 models were constructed using the logistic regression method, and then verified using the data from cohort 2 to calculate the accuracy and AUC values ​​of different models in cohort 2. Among them, the model constructed by cyclic adenosine monophosphate + creatine kinase isoenzyme was the best, with an accuracy of 67.39% and an AUC value of 0.803 in the validation cohort (cohort 2), which was better than other combinations. This shows that the data combination of Cyclic AMP+CKMB is effective for constructing a stable gout recurrence prediction model.

[0099] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit the scope of protection of the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application may be modified or replaced by equivalents without departing from the essence and scope of the technical solutions of the present application.

Claims

1. A biomarker for rheumatic gout, characterized in that: The gouty rheumatic disease biomarker includes at least one of acetylvanillone, cyclic adenosine monophosphate, methyl vanillate and uridine succinate.

2. The gouty rheumatism biomarker according to claim 1, characterized in that: The gouty rheumatic syndrome biomarkers include gouty rheumatic syndrome diagnostic markers and gouty rheumatic syndrome relapse prediction markers; The gouty rheumatic syndrome diagnostic marker includes at least one of acetylvanillone, cyclic adenosine monophosphate and methyl vanillate; The predictive markers for relapse of gouty rheumatic disease include cyclic adenosine monophosphate and uridine succinate.

3. The gouty rheumatism biomarker according to claim 2, characterized in that: The gouty rheumatic syndrome diagnostic markers are acetylvanillone and cyclic adenosine monophosphate.

4. The gouty rheumatism biomarker according to claim 2, characterized in that: The predictive markers for recurrence of gouty rheumatic disease also include clinical indicators.

5. The gouty rheumatism biomarker according to claim 4, characterized in that: The clinical indicators include creatine kinase isoenzyme or low-density lipoprotein cholesterol.

6. Use of the gouty rheumatic syndrome biomarker according to any one of claims 1 to 5 in a gouty rheumatic syndrome diagnosis product and / or a gouty rheumatic syndrome recurrence prediction product.

7. A kit for diagnosing or predicting recurrence of gout-rheumatic syndrome, characterized in that: The kit comprises a reagent for detecting the gouty rheumatic disease biomarker according to any one of claims 1 to 5.

8. A method for constructing a diagnostic model for gout and rheumatism, characterized in that: The following steps are involved: S1. Metabolite differences among healthy subjects, patients with rheumatic gout, and patients with non-rheumatic gout were analyzed, and the AUC values ​​of the metabolites were calculated to screen out acetylvanoic acid, cyclic adenosine monophosphate, and methyl vanillate. S2. Use the random forest algorithm to read data from healthy individuals, patients with rheumatic gout, and patients with non-rheumatic gout. Divide the data into a training set and a validation set. Use at least two of acetylvanoyl, cyclic adenosine monophosphate, and methyl vanillate as variables to set model parameters. After running the model, extract the out-of-bag error of the training set and the accuracy of the validation set. The combination of acetylvanoyl, cyclic adenosine monophosphate, and methyl vanillate was used as the variable, and the parameters of the model were set as follows: ntree = 2000, mtry = 2, proximity = TRUE, importance = TRUE, keep.inbag = TRUE; Using the combination of acetylvanillone and cyclic adenosine monophosphate as a variable, or the combination of acetylvanillone and methyl vanillate as a variable, or the combination of cyclic adenosine monophosphate and methyl vanillate as a variable, the model parameters are set as follows: ntree=2000, mtry=1, proximity=TRUE, importance=TRUE, keep.inbag=TRUE.

9. A method for constructing a recurrence prediction model for gout-rheumatic syndrome, characterized in that: The following steps are involved: S1. OPLS-DA algorithm was used to screen the differential metabolites between recurrent gouty rheumatic syndrome and non-recurrent gouty rheumatic syndrome according to the thresholds of VIP>1.5 and P<0.05 to obtain cyclic adenosine monophosphate and uridine succinate; S2. Different combinations of cyclic adenosine monophosphate, uridine succinate, and differential clinical indicators were used to build models using the logistic regression method. The models were then validated and the accuracy and AUC values ​​were calculated to evaluate the predictive performance of different combinations.

Images