Molecular marker for diagnosing necrotic soft tissue infection as well as application and model of molecular marker

Abstract

The invention belongs to the technical field of clinical examination and diagnosis, and relates to a molecular marker for diagnosing necrotic soft tissue infection as well as application and a model of the molecular marker. The molecular marker is selected from one or a combination of more of estrone sulfate, 3-hydroxydecanoic acid and N-lactoyl-phenylalanine. The molecular marker, the application and the construction method can be quickly, simply, sensitively and specifically used for early diagnosis of necrotic soft tissue infection.

Description

Technical Field

[0001] This invention belongs to the field of clinical laboratory diagnostic technology, and relates to molecular markers for the diagnosis of necrotizing soft tissue infections, their uses and models. Background Technology

[0002] Necrotizing soft tissue infections (NSTIs) are a rapidly progressing and fatal type of soft tissue infection. Clinically, NSTIs are mainly classified into two categories based on bacteriological differences: mixed infections with multiple microorganisms (Type I) and single-microorganism infections (Type II). Although the incidence of NSTIs is low (approximately 0.3 to 5 cases per 100,000 people per year), NSTIs have extremely high mortality and disability rates, causing severe physical and psychological harm to patients and placing a heavy burden on families and society.

[0003] Early diagnosis is a prerequisite for successful treatment of NSTI, but due to the atypical clinical symptoms of NSTI, the misdiagnosis rate is as high as 71%. Although imaging and the necrotizing fasciitis laboratory risk index (LRINEC) score have some value in the diagnosis of NSTI, they each have limitations in clinical application. To date, how to identify NSTI early and differentiate between necrotic and non-necrotic soft tissue infections remains a prominent challenge in clinical practice.

[0004] Therefore, finding rapid and sensitive molecular markers for the diagnosis of NSTI and developing an innovative, quick, and easy new diagnostic method for NSTI has significant clinical implications and socioeconomic benefits. Summary of the Invention

[0005] The primary objective of this invention is to provide molecular markers for the diagnosis of necrotizing soft tissue infections, enabling rapid, convenient, sensitive, and specific early diagnosis of these infections.

[0006] To achieve this objective, in a basic embodiment, the present invention provides a molecular marker for the diagnosis of necrotizing soft tissue infection, said molecular marker being selected from estrone sulfate (CAS number: 481-97-0 One or more of the following: 3-hydroxydecanoic acid (CAS number: 14292-26-3), N-lactoyl-phenylalanine (CAS number: 183241-73-8).

[0007] The method for screening molecular markers for the diagnosis of necrotizing soft tissue infections according to the present invention includes the following steps:

[0008] (1) Plasma samples were collected from patients with different necrotizing soft tissue infections and patients without necrotizing soft tissue infections, and the combinations of plasma samples from patients with different necrotizing soft tissue infections and patients without necrotizing soft tissue infections were used as analytical samples for the development cohort and the validation cohort, respectively.

[0009] (2) Perform non-targeted metabolomics analysis on each sample to obtain the original metabolic fingerprint of each plasma sample;

[0010] (3) The original metabolic fingerprint profiles of each plasma sample were compared and analyzed to obtain the differential metabolites between patients with necrotizing soft tissue infection and patients without necrotizing soft tissue infection.

[0011] (4) Screen out differentially expressed metabolites that are upregulated in patients with necrotizing soft tissue infection compared to patients without necrotizing soft tissue infection, and use differentially expressed metabolites that may be related to skin and soft tissue infection to perform diagnostic modeling. Finally, obtain the molecular markers as described above that distinguish patients with necrotizing soft tissue infection from those without necrotizing soft tissue infection.

[0012] Preferably, in step (2), the non-targeted metabolomics analysis is a technique for comprehensively analyzing all metabolites in a biological sample, without pre-setting specific metabolites as analytical targets. Through this method, researchers can obtain extensive metabolic data without being limited to specific compounds, thereby helping to reveal complex metabolic processes within organisms. Non-targeted metabolomics can detect unknown metabolites, providing information for disease mechanism research, the discovery of new biomarkers, and the development of new drugs.

[0013] Preferably, in step (2), the non-targeted metabolomics analysis is performed using LC-MS.

[0014] Preferably, in step (3), the data comparison analysis includes data preprocessing, data quality control, and multivariate statistical analysis.

[0015] Preferably, the data preprocessing refers to a series of operations performed on the data before data analysis, with the aim of cleaning and transforming the raw data to improve data quality and ensure the accuracy of the analysis results. This typically includes steps such as noise removal, handling missing values, normalization, and data normalization.

[0016] The data quality control described herein is a series of operations performed during data collection and analysis to ensure the accuracy and consistency of the data. This includes measures such as validating data acquisition methods, detecting outliers in the data, and comparing against reference standards to guarantee the quality and reliability of the final analysis.

[0017] Multivariate statistical analysis refers to the use of statistical methods to simultaneously analyze the relationships between multiple variables. This analysis can reveal whether there are correlations between variables, such as causal relationships or clustering relationships. Methods of multivariate statistical analysis include principal component analysis, factor analysis, cluster analysis, and discriminant analysis.

[0018] Preferably, in step (4), the diagnostic modeling adopts the random forest algorithm.

[0019] Preferably, the random forest algorithm randomly selects 70% as the training set and 30% as the validation set.

[0020] In a preferred embodiment, the present invention provides molecular markers for the diagnosis of necrotizing soft tissue infections, wherein the molecular markers are derived from plasma metabolites.

[0021] A second object of the present invention is to provide the use of the molecular markers described above for the preparation of reagents and / or kits for the diagnosis of necrotizing soft tissue infections, so as to enable rapid, simple, sensitive and specific early diagnosis of necrotizing soft tissue infections.

[0022] To achieve this objective, in a basic embodiment, the present invention provides the use of the molecular markers described above for the preparation of reagents and / or kits for diagnosing necrotizing soft tissue infections.

[0023] A third objective of this invention is to provide a method for constructing a diagnostic model for necrotizing soft tissue infection based on the molecular markers described above, so as to enable rapid, simple, sensitive, and specific early diagnosis of necrotizing soft tissue infection.

[0024] To achieve this objective, in a basic implementation scheme, the present invention provides a method for constructing a diagnostic model for necrotizing soft tissue infection based on the molecular markers described above, the method comprising the following steps:

[0025] (1) Plasma samples were collected from patients with different necrotizing soft tissue infections and patients without necrotizing soft tissue infections, and the combinations of plasma samples from patients with different necrotizing soft tissue infections and patients without necrotizing soft tissue infections were used as analytical samples for the development cohort and the validation cohort, respectively.

[0026] (2) Perform non-targeted metabolomics analysis on each sample to obtain the original metabolic fingerprint of each plasma sample;

[0027] (3) The original metabolic fingerprint profiles of each plasma sample were compared and analyzed to obtain the differential metabolites between patients with necrotizing soft tissue infection and patients without necrotizing soft tissue infection.

[0028] (4) Screen out differentially expressed metabolites that are upregulated in patients with necrotizing soft tissue infection compared to patients without necrotizing soft tissue infection, and use differentially expressed metabolites that may be related to skin and soft tissue infection to perform diagnostic modeling. Finally, obtain the molecular markers as described above that distinguish patients with necrotizing soft tissue infection from those without necrotizing soft tissue infection.

[0029] In a preferred embodiment, the present invention provides a method for constructing a diagnostic model for necrotizing soft tissue infection based on the molecular markers described above, wherein in step (2), the non-targeted metabolomics analysis is a technique for comprehensively analyzing all metabolites in a biological sample, without pre-setting specific metabolites as analytical targets. Through this method, researchers can obtain extensive metabolic data without being limited to specific compounds, thereby helping to reveal complex metabolic processes within organisms. Non-targeted metabolomics can detect unknown metabolites, providing information for disease mechanism research, the discovery of new biomarkers, and the development of new drugs.

[0030] In a preferred embodiment, the present invention provides a method for constructing a diagnostic model for necrotizing soft tissue infection based on molecular markers as described above, wherein in step (2), the non-targeted metabolomics analysis employs LC-MS coupled technology.

[0031] In a preferred embodiment, the present invention provides a method for constructing a diagnostic model for necrotizing soft tissue infection based on molecular markers as described above, wherein in step (3), the data comparison analysis includes data preprocessing, data quality control, and multivariate statistical analysis.

[0032] In a preferred embodiment, the present invention provides a method for constructing a diagnostic model for necrotizing soft tissue infection based on the molecular markers described above, wherein:

[0033] The data preprocessing described above refers to a series of operations performed on the data before data analysis. The purpose is to clean and transform the raw data to improve its quality and ensure the accuracy of the analysis results. This typically includes steps such as noise removal, handling missing values, normalization, and data normalization.

[0034] The data quality control described herein is a series of operations performed during data collection and analysis to ensure the accuracy and consistency of the data. This includes measures such as validating data acquisition methods, detecting outliers in the data, and comparing against reference standards to guarantee the quality and reliability of the final analysis.

[0035] Multivariate statistical analysis refers to the use of statistical methods to simultaneously analyze the relationships between multiple variables. This analysis can reveal whether there are correlations between variables, such as causal relationships or clustering relationships. Methods of multivariate statistical analysis include principal component analysis, factor analysis, cluster analysis, and discriminant analysis.

[0036] In a preferred embodiment, the present invention provides a method for constructing a diagnostic model for necrotizing soft tissue infection based on molecular markers as described above, wherein in step (4), the diagnostic modeling adopts a random forest algorithm.

[0037] In a preferred embodiment, the present invention provides a method for constructing a diagnostic model for necrotizing soft tissue infection based on molecular markers as described above, wherein the random forest algorithm randomly selects 70% as the training set and 30% as the validation set.

[0038] The beneficial effect of this invention is that, by utilizing the molecular markers, uses, and construction methods of this invention, it is possible to quickly, easily, sensitively, and specifically diagnose necrotizing soft tissue infections in their early stages.

[0039] The molecular markers of this invention have good sensitivity and specificity for the diagnosis of NSTI, and also have good sensitivity and specificity for distinguishing between necrotizing soft tissue infection and non-necrotizing soft tissue infection. They can be used for non-invasive diagnosis of NSTI, which is of great significance for timely treatment after a clear diagnosis, improving the clinical prognosis of patients, and improving their survival rate and quality of life.

[0040] This invention analyzes plasma samples from a development cohort (12 non-NSTI patients and 28 NSTI-infected patients) and a validation cohort (14 non-NSTI patients and 18 NSTI-infected patients). Fingerprints of small molecule metabolites in both positive and negative ion modes are obtained using liquid chromatography-mass spectrometry (LC-MS). Differential metabolites are screened through univariate and multivariate analyses of the fingerprints of small molecule metabolites from both non-NSTI and NSTI patients. Differential metabolites that are upregulated in both the development and validation cohorts and are associated with skin and soft tissue are then used as diagnostic molecular markers for NSTI. These molecular markers have high clinical application and promotion value.

[0041] This invention employs plasma non-targeted metabolomics combined with machine learning to analyze metabolic data from both non-NSTI and NSTI patients, resulting in diagnostic molecular biomarkers and models suitable for NSTI patients. The diagnostic molecular biomarker screening method of this invention is highly operable, the model construction method is simple, and the resulting diagnostic model exhibits good performance, high sensitivity, and good specificity, making it suitable for the diagnosis of NSTI. Furthermore, this invention achieves diagnosis solely through blood sampling, eliminating the need for additional tissue sample culture or invasive surgical examinations. The molecular biomarkers involved in this invention effectively classify the metabolomics data of NSTI and non-NSTI patients, accurately distinguishing between them, and possess significant clinical application and promotional value. Attached Figure Description

[0042] Figure 1 For example, the queue developed in Implementation 1 ( Figure 1 A) and verification queue ( Figure 1 B) Partial least squares discriminant analysis (PLS-DA) of differentially expressed metabolites, such as Figure 1 A, Figure 1 As shown in B, R in PLS-DA for the development queue and verification queue. 2 The values ​​were 0.86 and 0.91, respectively, indicating that the metabolites annotated in the plasma had a good ability to distinguish between NSTI patients and non-NSTI patients.

[0043] Figure 2 For example, the queue developed in Implementation 1 ( Figure 2 A, Figure 2 B) and verification queue ( Figure 2 C Figure 2 D) ROC curve and Geni index of the random forest diagnostic model. Detailed Implementation

[0044] To better understand the technical solutions and advantages of the present invention, the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0045] Example 1: Screening of molecular biomarkers for NSTI diagnosis

[0046] (I) Research Subjects

[0047] This study included 72 plasma samples from the First Affiliated Hospital of Wenzhou Medical University, serving as both the development cohort and the validation cohort. The development cohort consisted of 12 non-NSTI patients and 28 NSTI patients recruited between November 2015 and April 2021; the validation cohort consisted of 14 non-NSTI patients and 18 NSTI patients recruited between May 2021 and December 2022.

[0048] (II) Plasma non-targeted metabolomics analysis using liquid chromatography-mass spectrometry (LC-MS / MS)

[0049] All plasma samples were centrifuged and stored at -80°C. During the study, plasma samples were retrieved, pretreated, and then subjected to untargeted metabolomics analysis using high-performance liquid chromatography-mass spectrometry (HPLC-MS / MS) to obtain raw metabolic fingerprints containing chromatographic and mass spectrometric information. The specific procedures are as follows:

[0050] 1. Instruments and reagents

[0051] The main experimental instruments include: a mass spectrometer (Q Exactive). TM HF (Thermo Fisher), chromatograph (Vanquish UHPLC, Thermo Fisher), chromatographic column (Hypesil Gold column, Thermo Fisher), low-temperature centrifuge (D3024R, Scilogex).

[0052] The main experimental consumables include: Waters Xselect CSH C18 column (100×4.6mm, 3.5μm), 1.5ml injection vial, 300μl inner tube, etc.

[0053] The main experimental reagents include: methanol (Thermo Fisher, mass spectrometry grade), ammonium acetate (Thermo Fisher, mass spectrometry grade), formic acid (Thermo Fisher), etc.

[0054] 2. Plasma sample pretreatment

[0055] Before plasma sample pretreatment, prepare 6 quality control samples (QC samples, which are made by mixing equal volumes of plasma samples). All non-NSTI and NSTI plasma samples, along with the quality control samples, are pretreated separately as follows:

[0056] (1) Take 100 μL of sample and place it in an EP tube, then add 400 μL of 80 wt% methanol aqueous solution;

[0057] (2) Vortex oscillation, let stand in ice bath for 5 min, centrifuge at 15000g and 4℃ for 20 min;

[0058] (3) Take a certain amount of supernatant and dilute it with mass spectrometry grade water until the methanol content is 53 wt%;

[0059] (4) Centrifuge at 15000g and 4℃ for 20min, collect the supernatant, and analyze it by LC-MS.

[0060] 3. Plasma non-targeted metabolomics detection

[0061] All pretreated plasma samples from both non-NSTI and NSTI patients were used as analytical samples. The samples were randomly ordered and injected after being shuffled to eliminate bias from the injection order. A quality control sample was added every 10 analytical samples. The liquid chromatography and mass spectrometry methods used are as follows:

[0062] (1) Chromatographic conditions

[0063] Column: Hypesil Goldcolumn (C18)

[0064] Column temperature: 40℃

[0065] Flow rate: 0.2 mL / min

[0066] Positive mode: Mobile phase A: 0.1 wt% formic acid; Mobile phase B: methanol

[0067] Negative mode: Mobile phase A: 5 mM ammonium acetate, pH 9.0; Mobile phase B: methanol

[0068] Chromatographic gradient elution program:

[0069]

[0070] (2) Mass spectrometry conditions: Scan range selected m / z 100-1500; ESI source settings are as follows: spray voltage: 3.2kV; shear gas flow rate: 40arb; auxiliary gas flow rate: 10arb; capillary temperature: 320℃; polarity: positive; negative; MS / MS secondary scan mode: data dependent scan.

[0071] (III) Screening of plasma metabolic biomarkers

[0072] The samples were analyzed under the chromatographic and mass spectrometric conditions described above to obtain the raw metabolic fingerprints of all samples. After data preprocessing and metabolite identification, diagnostic molecular markers that can distinguish between non-NSTI and NSTI were screened. The specific procedures are as follows:

[0073] 1. Data preprocessing and metabolite identification

[0074] The raw data file was imported into the CD 3.1 database search software for processing. Simple screening of parameters such as retention time and mass-to-charge ratio was performed for each metabolite. Then, a retention time deviation of 0.2 min and a mass deviation of 5 ppm were set, and peak alignment was performed on different samples to improve identification accuracy. Subsequently, peak extraction was performed using information such as a mass deviation of 5 ppm, a signal intensity deviation of 30%, a signal-to-noise ratio of 3, a minimum signal intensity, and added ions. Peak area was also quantified. Target ions were then integrated, and molecular formulas were predicted using molecular ion peaks and fragment ions, and compared with the mzCloud (https: / / www.mzcloud.org / ), mzVault, and Masslist databases. Background ions were removed using blank samples, and the original quantitative results were standardized. Finally, the identification and relative quantification results of the metabolites were obtained.

[0075] 2. Data quality control

[0076] When performing LC-MS plasma non-targeted metabolomics analysis, the prepared QC samples were evenly inserted into the analytical samples in a sequence of one QC sample for every 10 analytical samples. This was used to monitor the quality control of the analytical samples from sample pretreatment to analysis in real time. After the raw metabolic fingerprint was preprocessed by CD 3.1 library search software, the coefficient of variation (%RSD) of each metabolite in the QC sample was calculated. The coefficient of variation of most metabolites was controlled below 30%, indicating that the quality control of the samples from sample pretreatment to analysis was good, and the obtained metabolomics data were true and reliable.

[0077] 3. Data processing and analysis

[0078] The identified metabolites were annotated using the KEGG database (https: / / www.genome.jp / kegg / pathway.html), the HMDB database (https: / / hmdb.ca / metabolites), and the LIPIDMaps database (http: / / www.lipidmaps.org / ). For multivariate statistical analysis, the data was transformed using the metabolomics data processing software MetaX, followed by partial least squares discriminant analysis (PLS-DA) to obtain the variable projection importance (VIP) value for the first principal component of each metabolite. For univariate analysis, t-tests were used to calculate the statistical significance (P-value) of each metabolite between the two groups, and the fold change (FC) of the metabolite between the two groups was also calculated.

[0079] 4. Screening for differentially expressed metabolites

[0080] Based on the default screening criteria for differentially expressed metabolites (VIP > 1.0, P < 0.05, and FC > 1.5 or FC < 0.667), 91 differentially expressed metabolites were selected from the development cohort sample, and 365 differentially expressed metabolites were selected from the validation cohort sample. The data were transformed using the metabolomics data processing software MetaX, and then partial least squares discriminant analysis (PLS-DA) was performed to obtain the variable projected importance for each metabolite. The VIP value represents the contribution of the metabolite to the grouping. In the PLS-DA model, R... 2 The closer the value is to 1, the more stable and reliable the model is. For example... Figure 1 A, Figure 1 As shown in B, R in PLS-DA for the development queue and verification queue. 2 The values ​​were 0.86 and 0.91, respectively, indicating that the metabolites annotated in the plasma had a good ability to distinguish between NSTI patients and non-NSTI patients.

[0081] 5. Screening of candidate biomarkers

[0082] A total of 22 differentially expressed metabolites were identified at the intersection of the development and validation cohorts, of which 10 were upregulated (see Table 1). Literature review revealed three differentially expressed metabolites associated with skin and soft tissue infections: estronesulfate, 3-hydroxydecanoic acid, and N-lactoyl-phenylalanine.

[0083] Table 1. 10 candidate plasma differential metabolites

[0084]

[0085] 6. Constructing the NSTI diagnostic model using the random forest algorithm.

[0086] Random forest modeling was performed using the R language's randomForest package for three metabolites: estrone sulfate, 3-hydroxydecanoic acid, and N-lactyl-phenylalanine. The Geni index was used to evaluate the contribution of each metabolite to the model. Results showed that the model effectively distinguished between non-NSTI and NSTI cases. The area under the receiver operating characteristic (ROC) curve (AUC) for the development and validation cohorts was 0.93 (…). Figure 2 A) and 0.99 ( Figure 2C). A review of published literature indicates that these three plasma metabolic markers were all discovered for the first time in the diagnosis of NSTIs, possessing significant clinical importance. Among them, estrone sulfate, 3-hydroxydecanoic acid, and N-lactyl-phenylalanine all exhibited high Gini indices in the development and validation cohorts. Figure 2 B. Figure 2 D), therefore it may be an important plasma diagnostic marker for NSTI.

[0087] It is generally believed that a diagnostic method with an AUC > 0.7 has good diagnostic efficacy, while the diagnostic model constructed in this invention achieves an AUC value of over 0.90. Therefore, the diagnostic biomarkers of this invention can effectively distinguish between non-NSTIs and NSTIs, which is beneficial for the clinical auxiliary diagnosis of non-NSTIs and NSTIs. This greatly helps to improve the clinical diagnosis, treatment, and assessment of NSTIs, and has good clinical application and promotion value.

[0088] In practical applications, more samples can be selected for modeling according to the modeling method of this invention to increase the accuracy of the model.

[0089] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims and their equivalents, this invention is also intended to include these modifications and variations. The above embodiments or implementations are merely illustrative examples of this invention, and it can also be implemented in other specific ways or forms without departing from its gist or essential characteristics. Therefore, the described embodiments should be considered illustrative rather than limiting in any respect. The scope of this invention should be defined by the appended claims, and any changes equivalent to the intent and scope of the claims should also be included within the scope of this invention.

Claims

1. Molecular marker for the diagnosis of necrotic soft tissue infections, characterized in that: The molecular marker is selected from the group consisting of one or more of the following: estrone sulfate, 3-hydroxydecanoic acid, N-lactoyl-phenylalanine.

2. The molecular marker of claim 1, characterized in that: The molecular marker is derived from plasma metabolites.

3. Use of the molecular marker of claim 1 or 2 for preparing a reagent and / or a kit for diagnosing necrotizing soft tissue infection.

4. A method for constructing a diagnostic model for necrotic soft tissue infection based on the molecular marker according to claim 1 or 2, characterized by, The construction method comprises the following steps: (1) Collecting plasma samples of different necrotizing soft tissue infection patients and non-necrotizing soft tissue infection patients respectively, and using combinations of the plasma samples of different necrotizing soft tissue infection patients and non-necrotizing soft tissue infection patients as analysis samples of a development cohort and a validation cohort respectively; (2) Performing non-targeted metabolomics analysis on each analysis sample to obtain original metabolic fingerprinting of each plasma sample; (3) Performing data comparison analysis on the obtained original metabolic fingerprinting of each plasma sample to obtain differential metabolites of necrotizing soft tissue infection patients relative to non-necrotizing soft tissue infection patients; (4) Screening differential metabolites that are up-regulated in necrotizing soft tissue infection patients relative to non-necrotizing soft tissue infection patients, performing diagnostic modeling on differential metabolites that are possibly related to skin soft tissue infection, and finally obtaining the molecular marker of claim 1 or 2 that distinguishes necrotizing soft tissue infection patients from non-necrotizing soft tissue infection patients.

5. The method of construction of claim 4, wherein: In step (2), the non-targeted metabolomics analysis uses LC-MS technology.

6. The method of construction of claim 4, wherein: In step (3), the data comparison analysis includes data preprocessing, data quality control, and multivariate statistical analysis.

7. The construction method of claim 6, wherein: The data preprocessing is a series of operations performed on data before data analysis, aiming to clean and transform original data to improve data quality and ensure the accuracy of analysis results, including one or more of the following operations: removing noise, processing missing values, standardization, and data normalization; The data quality control is a series of operations performed during data collection and analysis to ensure data accuracy and consistency, including one or more of the following operations: verifying data acquisition methods, detecting outliers in data, and comparing reference standards to ensure the quality and reliability of the final analysis; The multivariate statistical analysis refers to simultaneously analyzing the relationship between multiple variables using statistical methods to reveal whether there is a correlation between variables.

8. The construction method of claim 4, wherein: In step (4), the diagnostic modeling uses a random forest algorithm.

9. The method of construction of claim 8, wherein: The random forest algorithm randomly selects 70% as the training set and 30% as the validation set.

Images