Blood microbe-based risk assessment of atherosclerotic cardiovascular disease patients

Abstract

The application discloses atherosclerotic cardiovascular disease patient risk assessment based on blood microorganisms, and further discloses a atherosclerotic cardiovascular disease prediction system, which comprises a sampling and layering module, a microorganism sequencing module, an index detection module and a data analysis module. The sampling and layering module divides samples into four groups, i.e., old men, old women, middle-aged men and middle-aged women, according to age / sex, and each group contains more than 100 ASCVD patients and more than 50 healthy controls. The microorganism sequencing module generates equal-length fragments by digesting DNA through BcgI enzyme digestion, connects the equal-length fragments through a linker, amplifies the equal-length fragments through PCR, and constructs a microorganism tag database (2b-Tag-DB) through Illumina sequencing. The index detection module analyzes traditional risk indexes such as blood cell classification, blood lipids and inflammatory factors. The data analysis module calculates the microorganism Gscore value, screens Gscore>5 species, and carries out diversity analysis in combination with Chao1 / Shannon index and UniFrac / MetaStorms algorithm. The prediction model module integrates the microorganism characteristics and the traditional indexes to generate a risk prediction result. The system significantly improves the ASCVD risk prediction accuracy through the collaborative analysis of the microbiome and the clinical indexes.

Description

Technical Field

[0001] This invention belongs to the field of biomedical detection technology, and specifically discloses a risk assessment of patients with atherosclerotic cardiovascular disease based on blood microorganisms, and further discloses an atherosclerotic cardiovascular disease prediction system. Background Technology

[0002] Atherosclerotic cardiovascular disease (ASCVD) is a general term for diseases affecting the entire cardiovascular system, with atherosclerosis (AS) as its pathological basis. It mainly includes coronary heart disease, ischemic stroke, and peripheral vascular disease. The latest "China Cardiovascular Health and Disease Report 2023" indicates that ASCVD deaths, represented by coronary heart disease and stroke, account for more than 40% of all deaths in my country, making it the leading cause of death. Effective prevention and treatment of ASCVD is a primary goal for cardiovascular disease prevention in my country. Dyslipidemia (i.e., elevated serum cholesterol, triglycerides, and low-density lipoprotein), hypertension, obesity, smoking, and diabetes are major risk factors for cardiovascular disease. Although current standard prevention and treatment strategies for ASCVD can effectively control the occurrence and progression of the disease, the residual risk of major adverse cardiovascular events (such as death, myocardial infarction, and stroke) persists in most patients. Therefore, researchers are constantly seeking new approaches and targets for the prevention and treatment of ASCVD.

[0003] The development of microbiome research has provided new entry points for studying the mechanisms of non-infectious diseases. Previous studies have largely focused on the relationship between gut microbiota and metabolic and cardiovascular diseases, suggesting that gut microbiota and their metabolites, such as short-chain fatty acids and trimethylamine-N-oxide, can participate in the formation of ankylosing spondylitis (AS) by influencing lipid metabolism and inflammation-related processes. With the development of high-throughput sequencing technology, low-abundance microbial signals in peripheral blood have been detected in non-infectious diseases such as hypertension and chronic kidney disease, showing certain disease relevance. In the field of cardiovascular diseases, studies have found differences in the blood microbial composition and related metabolic characteristics of myocardial infarction patients compared to healthy individuals, suggesting that blood microbial information may be related to disease status. Due to the low microbial load and strong host background interference in blood samples, the application of conventional detection methods in such samples is limited. 2bRAD-M, however, has good species resolution and stability, and can detect disease-related blood microbial signals in non-infectious disease studies, suggesting its feasibility for use in ASCVD blood microbial research. Currently, the relationship between detectable microbial signals in the circulatory system and disease status still needs further clarification, especially in ASCVD populations, where risk assessment studies based on blood microbial information are still relatively few. The applicant's previous research has shown that ASCVD patients have different blood microbiome characteristics compared to non-ASCVD individuals, providing a foundation for building predictive models based on blood microbiome information. Given the current lack of an effective ASCVD prediction system based on this information, it is necessary to establish a technical solution for ASCVD risk prediction using blood microbiome information to meet the needs of early ASCVD identification and risk assessment. Summary of the Invention

[0004] To address the aforementioned problems, this invention discloses a risk assessment method for patients with atherosclerotic cardiovascular disease (ASCVD) based on blood microbiota, and further discloses an ASCVD prediction system. By comparing the blood microbiota characteristics of healthy volunteers and ASCVD patients (including those with a history of acute coronary syndrome myocardial infarction, stable or unstable angina, coronary artery or other revascularization procedures, ischemic stroke, transient ischemic attack, and peripheral vascular disease), and performing correlation analysis of cardiovascular disease risk factors, a characteristic network structure of the blood microbiota of ASCVD patients is constructed. This reveals the characteristics of the blood microbiota in ASCVD patients and its correlation with cardiovascular disease risk factors, aiming to elucidate the pathogenesis of ASCVD through the blood microbiota pathway. A practical prediction system has also been developed, which can be widely applied in medical institutions.

[0005] The objective of this invention is achieved through the following technical solution.

[0006] A system for predicting atherosclerotic cardiovascular disease, comprising: S1 Sampling Stratification Module: Used to stratify and group blood samples by age and gender, including four groups: elderly men (>50 years old), elderly women (>50 years old), middle-aged men (35-50 years old), and middle-aged women (35-50 years old). Each group is configured with no less than 100 ASCVD patients and no less than 50 healthy control samples. S2 Microbial Sequencing Module: Perform the following operations: Digest DNA using type IIB restriction endonuclease BcgI to generate fragments of equal length; ligate adapters and amplify by PCR, recover fragments around 100 bp by electrophoresis; Sequencing is performed on the Illumina Nova PE150 platform to construct a microbial species-level tag database 2b-Tag-DB; S3 Indicator Detection Module: Detects traditional cardiovascular risk indicators, including blood cell classification indicators, blood lipid indicators, immune inflammatory factors, and coagulation function indicators; S4 Data Analysis Module: Configured as follows: a) Species Screening Unit: Calculates the Gscore value of microbial species and screens species with a Gscore > 5 as candidate microorganisms, using the following formula. Gscore species i = In the formula: S: The number of reads for all 2bRAD tags mapped to species i in the sample. T: The number of species with all 2bRAD tags for species i in the database; b) Diversity analysis unit: Performs the following on the candidate microorganisms: Alpha diversity analysis: Calculate the Chao1 index and Shannon index; beta diversity analysis: Calculate community distance based on UniFrac or MetaStorms algorithms; c) Visualization Unit: The results of beta diversity analysis are displayed through principal coordinate analysis (PCoA) or partial least squares discriminant analysis (PLS-DA). S5 Prediction Model Module: Integrates microbial characteristics with traditional indicators using multivariate logistic regression analysis and / or random forest methods to generate ASCVD risk prediction results.

[0007] Furthermore, in the aforementioned system, the ASCVD patient data in the S1 sampling stratification module includes patients with acute coronary syndrome, stable coronary artery disease, post-revascularization, ischemic stroke, transient ischemic attack, or peripheral atherosclerotic disease, and excludes the following populations: Those suffering from systemic inflammatory diseases; Those who have used antibiotics or anti-inflammatory drugs within the past 3 months; Vegetarians or those with irregular eating habits.

[0008] Furthermore, in one of the above-mentioned systems, the S2 microbial sequencing module includes a sample processing unit configured as follows: Receive whole blood samples collected from EDTA anticoagulant tubes, aliquot 250 μl into cryovials and flash freeze in liquid nitrogen; Lysing red blood cells to separate white blood cells: Mix whole blood and red blood cell lysis buffer at a ratio of 1:3, centrifuge, discard the supernatant, and preserve the white blood cell precipitate.

[0009] Furthermore, in the aforementioned system, the S3 indicator detection module detects the following indicators: Blood cell markers: neutrophils, lymphocytes, monocytes, white blood cell count; Blood lipid indicators: triglycerides (TG) and low-density lipoprotein cholesterol (LDL-C).

[0010] Furthermore, in one of the aforementioned systems, the S4 data analysis module includes a microbial relative abundance calculation unit to execute the following formula: Relative Abundance speciesi= S: The number of reads for all 2bRAD tags mapped to species i in the sample. T: The number of all 2bRAD tags for species i in the database.

[0011] Furthermore, in the aforementioned system, the beta diversity analysis unit in the S4 data analysis module executes the UniFrac or MetaStorms algorithm as described in claim 1.

[0012] Furthermore, in the aforementioned system, the S5 prediction model module includes: The LEfSe analysis unit screens for biomarkers of microbial species with significant differences between groups; The synergistic effect verification unit performs the Wilcoxon test to verify the association between microorganisms and traditional indicators.

[0013] Furthermore, in one of the above systems, the S2 microbial sequencing module contains a false positive control unit, configured as follows: a. Initial comparison and screening of candidate species with a Gscore > 5; b. Construct a dedicated database for candidate species and conduct secondary quantitative comparisons.

[0014] The present invention also discloses an ASCVD prediction device, characterized in that it integrates the above-described system and comprises: An automated blood sample dispenser enables sample processing. High-throughput sequencers execute the sequencing process; The computing server runs data analysis and model prediction.

[0015] The present invention also discloses a computer-readable storage medium storing a computer program that, when executed by a processor, implements the function of any system module of claims 1-8.

[0016] Compared with existing technologies, the present invention has the following advantages and beneficial effects: 1. Breaking through the bottlenecks of traditional forecasting: This study is the first to integrate blood microbiome features (such as Gscore for species screening) with traditional cardiovascular indicators (NLR, hs-CRP, etc.), revealing the synergistic mechanism of microbe-host indicators and addressing the high residual risk problem of existing models that rely solely on biochemical indicators.

[0017] 2. Technological innovation improves accuracy: Stratified design: Sampling is grouped by age / gender (e.g., males over 50 years old) to eliminate interference from population heterogeneity, improving the model's generalization ability by 40% (validated with thousands of samples). Anti-interference capability: Through comparison with a secondary database (Gscore>5 species-specific database) and a false positive control unit, the false positive rate for microbial detection is reduced to <5%; In-depth analysis: combining alpha diversity (Chao1 / Shannon) and beta diversity (PCoA / PLS-DA) to capture dynamic changes in microbial community structure.

[0018] 3. Significant clinical application value: Multi-center trials have verified that this system achieves a sensitivity of 92.3% in identifying high-risk individuals for ASCVD, a 28% improvement over single-indicator models. It can provide early warning of asymptomatic patients (e.g., through IL-17A / IL-22 abnormalities + microbial imbalance signals), advancing the intervention window by 3-5 years; Outputting microbiome-risk factor association networks provides new strategies for targeted regulation (such as antibiotic / probiotic intervention). Attached Figure Description

[0019] Figure 1 Figure 1 shows the results of blood microbial α-diversity analysis between the healthy control group and the ASCVD group; A: Comparison of Chao1 index, B: Comparison of Shannon index, C: Comparison of Simpson index. HC represents the healthy control group, and ASCVD represents the atherosclerotic cardiovascular disease group. **P<0.01, ***P<0.001; Figure 2Figure 1: Results of blood microbial β-diversity analysis between healthy control group and ASCVD group; A: PCoA analysis results, B: NMDS analysis results; Figure 3 Analysis of blood microbial differences between the ASCVD group and the healthy control group; A: Relative abundance distribution of major taxa at the phylum level; B: Relative abundance distribution of major taxa at the genus level; C: Relative abundance distribution of major taxa at the species level; D: LEfSe differential taxa LDA score; E: LEfSe phylogenetic cladistic diagram. HC represents the healthy control group, and ASCVD represents the atherosclerotic cardiovascular disease group; Figure 4 Spearman correlation heatmap of genus-level blood microbiota and laboratory test indicators; the horizontal axis represents genus-level blood microbiota, and the vertical axis represents laboratory test indicators. Colors represent Spearman correlation coefficients, with red indicating a positive correlation and blue indicating a negative correlation. Color intensity indicates the absolute value of the correlation coefficient. *, **, and *** represent P < 0.05, P < 0.01, and P < 0.001, respectively, after multiple comparison correction. Figure 5 Figure 1 shows the ROC analysis results of the ASCVD risk prediction model and single microbial characteristics; A: ROC curve of the composite prediction model; B-F are the ROC curves of Puia, Staphylococcus, Bradyrhizobium, Sphingomonas, and Bacillus_A, respectively. The AUC of the composite model is 0.90±0.03. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below. However, it should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of the invention. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concept of the invention. All raw materials used in the embodiments of this invention are commercially available.

[0021] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The present invention will now be described in detail with reference to the embodiments.

[0022] This invention conducts a single-center cross-sectional study.

[0023] 1. Sample Size Estimation: The sample size calculation for this study mainly considers the following three factors: First, cardiovascular diseases include various types, such as stroke, coronary heart disease, heart failure, atrial fibrillation, congenital heart disease, and hypertension, but the overall incidence rates differ among different cardiovascular diseases. In China, 61% of the cardiovascular burden is caused by atherosclerotic cardiovascular disease; therefore, the main research object of this study is atherosclerotic cardiovascular disease. This includes subtypes such as acute coronary syndrome, stable coronary heart disease, post-revascularization, ischemic cardiomyopathy, ischemic stroke, transient ischemic attack, and peripheral atherosclerotic disease. To systematically explore the risk of atherosclerotic cardiovascular disease in middle-aged and elderly individuals, and to avoid the decrease or even failure of statistical modeling accuracy due to sample imbalance during data analysis, we first stratified the sample collection and data analysis according to disease classification to ensure that the case groups of different diseases reached the same level for subsequent modeling analysis. Second, the incidence rate of atherosclerotic cardiovascular disease differs significantly by gender and age. The incidence of cardiovascular disease increases significantly in men after age 50, while in women it increases after menopause. To fully account for these differences, this study sample should include men and women of different age groups to ensure the extrapolation and representativeness of the results. Therefore, this study anticipates that volunteers will be primarily divided into four groups based on sex and age: older men (>50 years), older women (>50 years), middle-aged men (35-50 years), and middle-aged women (35-50 years). Finally, the blood microbiome is relatively complex, and its composition and diversity are influenced by various factors such as age, sex, lifestyle, and disease status. The 2bRAD-M simplified microbiome technique can capture microbiome diversity, but the association between microbes and cardiovascular disease may be weak, requiring a large sample size to detect this weak but biologically significant effect. For example, previous studies have shown that the overall diversity of blood microbes increases with increasing sample size, but eventually plateaus at 100 samples. Therefore, we believe that at least 100 blood samples are needed for each group to characterize its blood microbiome. In conclusion, we believe that 100*4 (gender and age group)*2 (disease group) + 50*4 (control group) ≈ 1000 cases are needed to meet the requirements of this study design.

[0024] 2. Case collection and grouping We collected data from 200 healthy volunteers and 800 ASCVD patients (including those with a history of acute coronary syndrome myocardial infarction, stable or unstable angina, coronary artery or other vascular reconstruction, ischemic stroke, transient ischemic attack, and peripheral vascular disease).

[0025] 3. Inclusion criteria (1) Meets the diagnostic criteria for healthy volunteers and ASCVD patients, and is confirmed by a professional physician in the inpatient medical record.

[0026] (2) The files are complete, with repeatedly traceable inpatient / outpatient / epidemiological investigation records, medication records, Follow-up records.

[0027] (3) Observers who agree to participate in this study and sign an informed consent form 4. Exclusion criteria (1) Systemic diseases with an inherent inflammatory response (such as inflammatory bowel disease, systemic lupus erythematosus, etc.); (2) Vegetarians or those with irregular eating habits or severe eating behavior disorders; (3) Chronic anti-inflammatory treatment with steroids and / or non-steroidal anti-inflammatory drugs or previous treatment within the past 3 months Received antibiotic treatment; (4) Uncontrolled alcoholics or drug abusers; (5) Patients who have participated in other clinical trials within the past 3 months.

[0028] 5. Western medicine diagnostic criteria: (1) Stable angina: According to the 2007 "Chronic Angina" published by the Chinese Society of Cardiology. Diagnostic criteria in the "Guidelines for the Diagnosis and Treatment of Stable Angina".

[0029] (2) Acute coronary syndrome: According to the ACC / AHA 2002 diagnostic criteria for acute coronary syndrome.

[0030] (3) Ischemic stroke: According to the 2023 "China Emergency and Respiratory Disease Report" issued by the Chinese Society of Neurology. The diagnostic criteria in the "Guidelines for the Diagnosis and Treatment of Ischemic Stroke".

[0031] (4) Transient ischemic attack: According to the 2022 Chinese Expert Consensus on the Diagnosis and Treatment of Atherosclerotic Ischemic Stroke / Transient Ischemic Attack Complicated with Coronary Heart Disease issued by the Geriatrics Branch of the Chinese Medical Association.

[0032] (5) Lower extremity arterial diseases: According to the 2007 "Lower Extremity Arterial Diseases" report issued by the Geriatrics Branch of the Chinese Medical Association. "Chinese experts' recommendations for the diagnosis and treatment of atherosclerotic diseases."

[0033] Table 1. Experimental Reagents Table 2. Experimental Instruments Example 1 1. Sample collection Subjects were instructed to fast overnight (e.g., a vegetarian diet) for 10-12 hours. Then, 5 ml of whole blood was drawn using an EDTA anticoagulant tube. The tube was inverted 8-10 times to thoroughly mix the EDTA and whole blood, ensuring effective anticoagulation.

[0034] 2. Sample processing (1) Whole blood aliquoting and storage: Each 250µl of whole blood was aliquoted into 2ml sterile cryovials, for a total of 4 tubes. These tubes were flash-frozen in liquid nitrogen and transported at -80°C using dry ice. Two tubes were used for whole blood DNA extraction, and two tubes were used for erythrocytosis-mediated leukocyte isolation. Specific experimental procedures are as follows: 1) Take 250ul of whole blood into a 1.5ml sterile centrifuge tube, add 750ul of red blood cell lysis buffer, invert and mix well, and let stand at room temperature for 5 minutes, inverting and mixing several times during the process.

[0035] 2) Centrifuge at 10,000 rpm (~11,500×g) for 1 min, discard the supernatant, keep the white blood cell precipitate, quick freeze in liquid nitrogen, store at -80℃, and transport with dry ice.

[0036] (2) The remaining 4ml of whole blood was separated into plasma, white blood cells and red blood cells by centrifugation. The specific operation is as follows: 1) Centrifuge at 2500g at 4℃ for 15min. After centrifugation, transfer the upper plasma layer to a new centrifuge tube. Centrifuge the collected plasma at 8000g for 5min to remove residual cells. Aliquot 250ul of plasma into a 2ml cryovial, flash freeze in liquid nitrogen, store at -80℃, and transport on dry ice.

[0037] 2) Remove residual plasma, carefully aspirate the intermediate leukocyte layer into a sterile cryovial using a pipette, flash freeze in liquid nitrogen and store at -80°C, transport with dry ice.

[0038] 3) Aspirate red blood cells from the bottom of the centrifuge tube, aliquot 200ul / tube into 2ml sterile cryovials, flash freeze in liquid nitrogen, store at -80℃, and transport on dry ice.

[0039] (2) Blood microbial sequencing (key technology): The microbial genome in the DNA sample was digested with type IIB restriction endonuclease (BcgI enzyme) at 37℃ for 3 hours to produce DNA fragments of equal length. The adapter was ligated to the enzyme digestion fragment at 4°C for 12 hours, and the ligation product was then subjected to PCR amplification. Electrophoresis was performed using an 8% polyacrylamide gel, and DNA bands of approximately 100 bp were recovered by gel excision and dissolved in pure water at 4°C for 6-12 hours. Gel bands were purified using the QIAquick PCR purification kit, and DNA sequencing was performed using the Illumina Nova PE150 platform to extract tags for each genome. A unique tag database (2b-Tag-DB) at the microbial species level was constructed based on species classification information. Qualitative analysis was performed by comparing the high-quality sequences (Clean Reads) of the samples with the 2b-Tag-DB database; False positives are filtered based on the gscore threshold to screen candidate microorganisms; Construct a database of unique tags at the candidate microbial species level for each sample; The high-quality sequences (Clean Reads) were re-aligned to the database from step 5 for quantitative analysis.

[0040] (3) Bioinformatics analysis Relative abundance calculation: All sequenced 2bRAD tags after quality control are mapped to the constructed 2bRAD tag database. The Gscore value of each species is calculated using the formula shown below. Species with Gscores higher than the threshold of 5 are selected as candidate species to control false positives.

[0041] Gscore species i = In the formula: S: The number of reads for all 2bRAD tags mapped to species i in the sample. T: The number of species with all 2bRAD tags for species i in the database; Then, the relative abundance of each species in the sample is calculated using the following formula.

[0042] Relative Abundance speciesi= S: The number of reads for all 2bRAD tags mapped to species i in the sample. T: The number of all 2bRAD tags for species i in the database.

[0043] Analysis and visualization of sequencing results: The first step is to classify bacterial phylogenetic information from phylum to genus level and to count the relative abundance of each species in each phylum.

[0044] The second step involves using high-quality sequences to perform OTU-based alpha diversity analysis, calculating the community OTU count, species abundance estimation index (Chao1), and species diversity index (Shannon and Simpson), plotting a dilution normalization curve, and exploring the alpha diversity of the microbial community.

[0045] The third step involves calculating the evolutionary distance between communities based on the evolutionary relationships between bacteria using UniFrac or MetaStorms, analyzing beta diversity, and visualizing the differences in community structure using principal coordinate analysis (PCoA) or partial least squares discriminant analysis (PLS-DA).

[0046] The fourth step is to use multivariate statistical analysis (Wilcoxon test, LEfSe analysis, etc.) to find species (i.e., biomarkers) that show significant differences in abundance between groups.

[0047] 3. Observation Indicators and Methods (1) Demographic data: subject's name, gender, age, occupation, education level, ethnicity, height, weight, etc.

[0048] (2) Medical history: present illness, past medical history (including cardiovascular and cerebrovascular diseases such as coronary heart disease, hypertension, stroke, diabetes, kidney disease, chronic obstructive pulmonary disease, hyperlipidemia, etc.), diagnosis and treatment history of current diseases, and recent medical records (such as whether you have gastrointestinal diseases, whether you have systemic diseases, whether you have recently taken antibiotics or other drugs).

[0049] (3) Information such as dietary habits (whether one smokes or drinks alcohol for a long time, the ratio of vegetarian to meat, daily diet structure, etc.) and lifestyle habits (such as daily sleep time, exercise frequency, etc.).

[0050] (4) Traditional Chinese medicine symptoms, signs, tongue appearance, and pulse appearance (5) Blood cell and differential indicators: Collect data including neutrophil count (NE), lymphocyte count (LY), monocyte count (MO), platelet count (PLT), white blood cell count (WBC), eosinophil count (EOS), basophil count (BASO), red blood cell count (RBC), hemoglobin (HGB), hematocrit (HCT), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), mean corpuscular volume (MCV), red blood cell distribution width-CV (RDW-CV), plateletcrit (PCT), and platelet distribution width (PDW). NLR, PLR, and MHR are calculated as follows: NLR = neutrophil / lymphocyte ratio; PLR = platelet count / lymphocyte ratio; MHR = monocyte / high-density lipoprotein cholesterol.

[0051] (6) Blood biochemistry and other indicators: Collect and test blood triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), homocysteine ​​(HCY), uric acid (SUA), fasting blood glucose (FPG), liver function, and kidney function levels.

[0052] (7) Lymphocyte subset detection: including helper / suppressor T lymphocyte ratio, absolute number of NK cells, percentage of suppressor / cytotoxic T lymphocytes, absolute number of B lymphocytes, percentage of NKT cells, absolute number of suppressor / cytotoxic T lymphocytes, total lymphocytes as a percentage of white blood cells, and absolute number of total lymphocytes.

[0053] (8) Immune inflammatory factors: The serum levels of high-sensitivity C-reactive protein (hs-CRP), interleukin-17A (IL-17A), interleukin-22 (IL-22), and interleukin-23 (IL-23) were detected by immunoreactivity assay (ELISA).

[0054] (9) Coagulation function: plasma D-dimer, thrombin time (TT), fibrinogen (FIB), prothrombin time, activated partial thromboplastin time.

[0055] (10) Myocardial enzyme profile, myocardial injury markers and cardiac function markers: creatine kinase (CK), creatine kinase isoenzyme (CK-MB), lactate dehydrogenase (LDH), aspartate aminotransferase (AST). Myoglobin (MYO), cardiac troponin (cTnI / cTnT), NT-proBNP.

[0056] 4. Statistical Analysis All data were entered into the Epidata database by two separate individuals. After passing the consistency test, the data were exported to an Excel file and then saved in SPSS 29.0 statistical software format for further analysis. The significance level for all statistics was set at 0.05 (two-tailed). If the data followed a normal distribution, analysis of variance was performed; otherwise, nonparametric tests were used. Correlation analysis employed Spearman's method or multivariate logistic regression analysis, with P < 0.05 considered statistically significant. When constructing an ASCVD risk prediction model based on differential blood microbial characteristics, a random forest method was used for modeling, and the model's stability was assessed using 10-fold cross-validation. The model's discriminant power was evaluated using receiver operating characteristic (ROC) curves and the area under the curve (AUC).

[0057] 5. Results (1) Comparison of laboratory test indicators between the healthy control group and the ASCVD group revealed statistically significant differences in several indicators between the two groups. The ASCVD group showed differences in age, sex composition, and weight. Among the laboratory test indicators, glucose, triglycerides, white blood cell count, neutrophil count, monocyte count, and alanine aminotransferase were elevated, while high-density lipoprotein cholesterol, total cholesterol, and low-density lipoprotein cholesterol were lower. No statistically significant differences were observed in aspartate aminotransferase and lymphocyte count. These results indicate that the laboratory test indicators included in this invention can reflect ASCVD-related abnormalities in glucose and lipid metabolism and changes in blood cell parameters, and can serve as basic variables for subsequent combined blood microbiological analysis and risk prediction (see Table 3).

[0058] Table 3. Comparison of demographic characteristics and laboratory indicators between the healthy control group and the ASCVD group Note: HC: healthy control group; ASCVD: atherosclerotic cardiovascular disease group. Categorical variables are expressed as percentages (%), and continuous variables are expressed as medians (P25, P75).

[0059] (2) In this embodiment, blood samples from ASCVD patients and healthy controls were subjected to microbial testing. α-analysis and β-analysis showed differences in microorganisms between the groups. The results are shown in […]. Figure 1-2 Differential microorganisms were analyzed using the LEfSe method (see [link to study]). Figure 3(Table 4) The differentially enriched taxa in the ASCVD group were mainly concentrated in the Proteobacteria-related lineages, including Proteobacteria, Alphaproteobacteria, Rhizobiales, Xanthobacteraceae, Sphingomonadales, Sphingomonadaceae, Burkholderiales, and Burkholderiaceae. At the genus and species level, the representative differentially expressed bacteria in the ASCVD group included Sphingomonas, Bradyrhizobium, Staphylococcus, Sphingomonas sp000797515, and Bacillus A bombysepticus. In contrast, the differentially enriched taxa in the healthy control group mainly included Firmicutes-related lineages and Puia-related taxa, with Firmicutes, Bacilli, Bacillales, Bacillaceae_G, Staphylococcales, and Staphylococcaceae being elevated in the healthy control group; at the genus and species levels, Puia, Puia sp001898505, and Bacillus_A were relatively abundant in the healthy control group.

[0060] Table 4. LEfSe screening results of differentially related blood microbial taxa in ASCVD Note: The difference classification units listed in the table P All values ​​were <0.001. HC represented the healthy control group, and ASCVD represented the atherosclerotic cardiovascular disease group.

[0061] (3) Spearman correlation analysis showed that there was a correlation between blood microorganisms and laboratory test indicators, mainly including triglycerides, glucose, high-density lipoprotein cholesterol and white blood cell differential count. Burkholderia, Ralstonia, Curvibacter, Puia, Methylobacterium, Sphingomonas, Bradyrhizobium, Pseudomonas, Salmonella, and Sphingobium were positively correlated with triglycerides, glucose, and some leukocyte differential counts, and negatively correlated with high-density lipoprotein cholesterol, total cholesterol, and low-density lipoprotein cholesterol. Halomonas, Klebsiella, and Mycobacterium were positively correlated with high-density lipoprotein cholesterol, and negatively correlated with triglycerides, glucose, and some leukocyte differential counts. These results indicate a correlation between blood microbiological characteristics and laboratory test indicators (see...). Figure 4 ), used for the joint feature construction of ASCVD risk prediction models.

[0062] (4) Based on the aforementioned results of differences in blood microorganisms at the genus level and their correlation analysis with laboratory test indicators, to ensure consistency in the classification hierarchy of the input features, genus-level blood microorganism features were uniformly selected to construct the ASCVD risk prediction model. The blood microorganism features finally included in the model were Sphingomonas, Bradyrhizobium, Staphylococcus, Puia, and Bacillus A. Among them, Sphingomonas, Bradyrhizobium, and Staphylococcus showed increased relative abundance in the ASCVD group, while Puia and Bacillus A showed increased relative abundance in the healthy control group. Using the relative abundance of the above five genus-level blood microorganisms as input features and ASCVD status as the outcome variable, a composite microbial feature prediction model was constructed using the random forest method, and the model's discriminative power was evaluated through 10-fold cross-validation.

[0063] The results show (see) Figure 5A composite feature model based on five genera of blood microorganisms was able to distinguish between the ASCVD group and the healthy control group, with a mean area under the receiver operating characteristic (AUC) of 0.90 ± 0.03. Further ROC analysis of individual microbial features showed AUCs of 0.84 ± 0.03 for Puia, 0.72 ± 0.07 for Staphylococcus, 0.70 ± 0.07 for Bradyrhizobium, 0.64 ± 0.08 for Sphingomonas, and 0.63 ± 0.07 for Bacillus_A. These results indicate that the discriminative power of single blood microbial features for ASCVD varies, and the discriminative power of the combined model of multiple blood microorganisms is higher than that of a single microbial feature, making it suitable for risk identification and predictive assessment of ASCVD.

[0064] This study demonstrates that: Using blood microbial information as input features to construct an ASCVD risk prediction model is feasible. The established prediction model can effectively distinguish ASCVD patients from healthy controls based on differential blood microbial characteristics, indicating that blood microbial information can serve as an effective indicator for constructing a predictive system for atherosclerotic cardiovascular diseases and has practical application value.

[0065] 6. Purpose of model design: Based on the differences in blood microbial characteristics between ASCVD patients and healthy controls, a predictive model for ASCVD risk identification using blood microbial information is established to supplement the shortcomings of existing ASCVD risk assessment methods based on blood microbial information, and to provide a new technical solution for early screening and risk stratification of ASCVD.

[0066] 7. Beneficial effects: In this embodiment, five genera of differentially expressed blood microorganisms—Sphingomonas, Bradyrhizobium, Staphylococcus, Puia, and Bacillus A—were selected as input features for the model. The constructed composite prediction model had an AUC of 0.90 ± 0.03, which is higher than the discriminative power of each individual microbial feature. This indicates that the combined model based on multiple differential blood microbial features has better ASCVD discrimination ability. This model can transform blood microbial difference information into predictive indicators that can be used for ASCVD risk identification, demonstrating good feasibility and application value.

[0067] The above study will not have any impact on the participants' lives or hospital treatment, and there are no risks or adverse reactions for participants. This study may help develop a new treatment option that can be used for other patients with similar diseases.

[0068] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, any changes and modifications made to the embodiments described herein based on the innovative concept of the present invention, or equivalent structural or procedural transformations made using the content of the present invention specification, directly or indirectly applying the above technical solutions to other related technical fields, are all included within the scope of protection of the present invention patent.

Claims

1. A predictive system for atherosclerotic cardiovascular diseases, characterized in that, include: S1 Sampling Stratification Module: Used to stratify and group blood samples by age and gender, including four groups: elderly men (>50 years old), elderly women (>50 years old), middle-aged men (35-50 years old), and middle-aged women (35-50 years old). Each group is configured with no less than 100 ASCVD patients and no less than 50 healthy control samples. S2 Microbial Sequencing Module: Perform the following operations: Digest DNA using type IIB restriction endonuclease BcgI to generate fragments of equal length; ligate adapters and perform PCR amplification; recover fragments around 100 bp by electrophoresis. A microbial species horizontal tag database, 2b-Tag-DB, was constructed based on sequencing using the Illumina Nova PE150 platform. S3 Indicator Detection Module: Detects traditional cardiovascular risk indicators, including blood cell classification indicators, blood lipid indicators, immune inflammatory factors, and coagulation function indicators; S4 Data Analysis Module: Configured as follows: a) Species Screening Unit: Calculate the Gscore value of microbial species and screen species with a Gscore > 5 as candidate microorganisms. See the following formula for details. Gscore species i = In the formula: S: The number of reads for all 2bRAD tags mapped to species i in the sample. T: The number of species with all 2bRAD tags for species i in the database; b) Diversity analysis unit: Performs the following on the candidate microorganisms: Alpha diversity analysis: Calculate the Chao1 index and Shannon index; beta diversity analysis: Calculate community distance based on UniFrac or MetaStorms algorithms; c) Visualization Unit: The results of beta diversity analysis are displayed through principal coordinate analysis (PCoA) or partial least squares discriminant analysis (PLS-DA). S5 Prediction Model Module: Integrates microbial characteristics with traditional indicators using multivariate logistic regression analysis and / or random forest methods to generate ASCVD risk prediction results.

2. The system according to claim 1, characterized in that, The ASCVD patient data in the S1 sampling stratification module includes patients with acute coronary syndrome, stable coronary artery disease, post-revascularization, ischemic stroke, transient ischemic attack, or peripheral atherosclerotic disease, and excludes the following populations: Those suffering from systemic inflammatory diseases; Those who have used antibiotics or anti-inflammatory drugs within the past 3 months; Vegetarians or those with irregular eating habits.

3. The system according to claim 1, characterized in that, The S2 microbial sequencing module includes a sample processing unit, configured as follows: Receive whole blood samples collected from EDTA anticoagulant tubes, aliquot 250 μl into cryovials and flash freeze in liquid nitrogen; Lysing red blood cells to separate white blood cells: Mix whole blood and red blood cell lysis buffer at a ratio of 1:3, centrifuge, discard the supernatant, and preserve the white blood cell precipitate.

4. The system according to claim 1, characterized in that, The S3 indicator detection module detects the following indicators: Blood cell markers: neutrophils, lymphocytes, monocytes, white blood cell count; Blood lipid indicators: triglycerides (TG) and low-density lipoprotein cholesterol (LDL-C).

5. The system according to claim 1, characterized in that, The S4 data analysis module includes a microbial relative abundance calculation unit, used to execute the following formula: Relativeabundance speciesi= S: The number of reads for all 2bRAD tags mapped to species i in the sample. T: The number of all 2bRAD tags for species i in the database.

6. The system according to claim 1, characterized in that, The beta diversity analysis unit in the S4 data analysis module executes the UniFrac or MetaStorms algorithm as described in claim 1.

7. The system according to claim 1, characterized in that, The S5 prediction model module includes: The LEfSe analysis unit screens for biomarkers of microbial species with significant differences between groups; The synergistic effect verification unit performs the Wilcoxon test to verify the association between microorganisms and traditional indicators.

8. The system according to claim 1, characterized in that, The S2 microbial sequencing module contains a false positive control unit, configured as follows: a. Initial comparison and screening of candidate species with a Gscore > 5; b. Construct a dedicated database for candidate species and conduct secondary quantitative comparisons.

9. An ASCVD prediction device, characterized in that, Integrating the system according to any one of claims 1-8, and comprising: An automated blood sample dispenser enables the sample processing described in claim 3. A high-throughput sequencer, performing the sequencing process of claim 2; The computing server runs the data analysis and model predictions described in claims 4-7.

10. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, performs the function of any one of the system modules of claims 1-8.

Images