Protein joint detection-based viral ARDS prognosis evaluation system and method

Abstract

The invention discloses a viral ARDS prognosis evaluation system based on protein joint detection, and the system comprises a sample processing module which is used for carrying out serum standardization pretreatment; the protein detection module is used for performing iBAQ quantification of IL6ST / FOXO3 / TLR7 based on DIA mass spectrometry, defining a targeted therapy threshold value and realizing targeted quantification of the core protein; and the intelligent analysis module is used for realizing cross-age risk layering based on a multivariable logic regression model. B cell function states are reflected by combining serum IL6ST / FOXO3 / TLR7 protein expression levels, a multivariable logistic regression model is constructed, and unified risk stratification of patients of all ages from children to the elderly is achieved; by locking an IL6ST / FOXO3 / TLR7 pathway, a targeted therapy threshold is defined to directly guide targeted therapy, and immune intervention is facilitated.

Description

Technical Field

[0001] The present invention relates to a viral ARDS prognosis assessment system and method, and in particular to a viral ARDS prognosis assessment system and method based on protein combined detection. Background Art

[0002] The existing prognostic assessment system for viral acute respiratory distress syndrome (ARDS), based on the GSE283501 public dataset, relies on the combined prediction of inflammatory factors such as C-reactive protein and interleukin, and has the following shortcomings: First, the existing prognostic assessment system for viral ARDS has significant age limitations. According to the Berlin definition published in the Journal of the American Medical Association (JAMA, 2012) and the consensus criteria of the Journal of Pediatric Critical Care Medicine, when there is no statistical difference in disease severity, the mortality rate for pediatric ARDS is 27.3% and that for adults is 75%, significantly lower than that for adults. When optimized for adults, the AUC dropped to 0.58 (P=0.03) when applied to children, failing to explain the paradoxical association between high inflammation and low mortality in children. For example, IL-6 levels in patients with MIS-C can be three times that of adults, yet their mortality rate is 47.5% lower than that of adults. Second, there are no objective prognostic markers based on B cell immune function heterogeneity to date, and the pathway of immune aging is not involved at all, resulting in a lack of precise targeting basis for immune intervention strategies. Summary of the Invention

[0003] In order to address the shortcomings of the above technologies, the present invention provides a viral ARDS prognosis assessment system and method based on protein combined detection.

[0004] In order to solve the above technical problems, the technical solution adopted in the present invention is: a viral ARDS prognosis assessment system based on protein combined detection, including a sample processing module for pretreatment of serum standardization; a protein detection module for iBAQ quantification of IL6ST / FOXO3 / TLR7 based on DIA mass spectrometry, defining targeted treatment thresholds, and realizing targeted quantification of core proteins; and an intelligent analysis module for realizing cross-age risk stratification based on a multivariate logistic regression model.

[0005] Furthermore, the intelligent analysis module, including the immune aging pathway integration unit, is used to perform mechanism analysis and treatment response prediction to address the problem of cognitive bias in pathological mechanisms.

[0006] The prognostic assessment method for viral ARDS based on combined protein detection specifically includes the following steps: Step S1: receiving a patient's serum sample and introducing it into the sample processing module of the ARDS prognostic evaluation system through an automatic sample injector for serum standardization; Step S2: The serum sample enters the constant temperature metal bath in the protein detection module to achieve thermal denaturation and inactivation of the protein; Step S3: Entering the digestion reaction chamber; the digestion reaction chamber is a constant temperature oscillation chamber at 37°C, and trypsin digestion is performed inside the digestion reaction chamber; wherein the oscillation frequency is 120 rpm, the mass ratio of enzyme to protein is 1:25, and the reaction time is 12 hours; Step S4: Peptide separation by liquid chromatography based on the EASY nLC 1200 system; Step S5: performing DIA mass spectrometry detection based on QE HF-X; Step S6: iBAQ quantification of IL6ST / FOXO3 / TLR7 based on the iProteome cloud platform; Step S7: Calculate the cross-age risk score based on the multivariate logistic regression model in the intelligent analysis module, and make a risk judgment to output the risk level; the formula for executing the risk score in the multivariate logistic regression model is: ; in, is the weight coefficient of IL6ST; is the weight coefficient of FOXO3; is the weight coefficient of TLR7; for risk thresholds; for risk thresholds; for risk thresholds; Measure the IL6ST expression value for the patient. Measure the FOXO3 expression value for the patient. The TLR7 expression value was measured for the patients.

[0007] Furthermore, in step S4, the chromatographic column for liquid chromatography separation of peptides was ReproSil-Pur C18-AQ; the flow rate was 300 nL / min, and the gradient was 5-35% acetonitrile / 120 min.

[0008] Furthermore, in step S5, the scanning range of the DIA mass spectrometry detection based on QE HF-X is m / z 300-1400; the MS1 resolution is 60000, the MS2 resolution is 15000; and the collision energy is 27%.

[0009] Furthermore, in step S5, DIA mass spectrometry detection is performed, which specifically includes the following steps: Step A1: The digested peptides were concentrated and dried by SpeedVac vacuum centrifugation to obtain peptide samples; Step A2: The peptide samples are separated by a high performance liquid chromatography system; Step A3: The eluted peptides are peptide samples eluted after separation on the chromatographic column; the eluted peptides are ionized at a voltage of 2 kV to maintain the integrity of the peptides; and are then transferred to the Q Exactive HF-X quadrupole-orbitrap mass spectrometer for data-independent acquisition.

[0010] Furthermore, the peptide sample in step A2 is separated by a high performance liquid chromatography system, specifically comprising the following steps: Step A21: re-dissolving the peptide sample in a solvent; Step A22: Load the sample onto a self-packed trapping column to adsorb and retain the target peptide, allowing impurities to elute with the mobile phase; Step A23: Separation is performed through an analytical column.

[0011] Furthermore, step S6 performs iBAQ quantification of IL6ST / FOXO3 / TLR7, specifically comprising the following steps: Step B1: The raw mass spectrometry data were input into the iProteome cloud platform for processing, and the label-free iBAQ absolute quantification method was used to calculate the total protein ratio × 10 3 Normalize the expression levels, and fill missing values ​​with the minimum non-zero value of 1%. Only proteins with missing values ​​< 50% were screened for differential analysis, and the significantly differentially expressed proteins were determined by t-test after log2 transformation. Step B2: Complete functional enrichment analysis using Metascape; Step B3: Through weighted gene co-expression network WGCNA analysis, after variance stabilization and log2(x+1) transformation, a soft threshold β=2 was selected to construct a topological overlap matrix; and a dynamic cutting method was used to identify co-expression modules. The minimum number of modules in the core parameters of the co-expression module was 30. When the similarity between modules was ≥0.25, they were merged to reduce redundant modules; Step B4: Perform cell enrichment analysis using the xCell algorithm and calculate the ssGSEA enrichment score based on the cell signature gene set. ssGSEA can calculate the enrichment score of a specific gene set for a single sample and quantify the expression activity of the gene set in the sample; Step B5: Prognostic evaluation of FOXO3, IL6ST, and TLR7 proteins.

[0012] Furthermore, in step S7, cross-age risk scoring is performed based on the multivariate logistic regression model in the intelligent analysis module, which can realize a method for constructing a multivariate logistic regression model for cross-age risk scoring, specifically comprising the following steps: Step D1: Select training data as the training set and preprocess the training data. The preprocessing includes cleaning and standardization. Step D2: Select samples from the GSE225349 dataset as the external validation set, and perform data cleaning and standardization on the data in the external validation set;

[0013] Step D3: Input the pre-processed training data in the training set into the multivariate logistic regression model for model training and perform iterative optimization to minimize the difference between the probability value predicted by the model and the measured value, thereby obtaining the optimal model parameters; Step D4: Calculate the model evaluation indicator AUC.

[0014] Furthermore, the immune aging pathway integration unit conducts mechanism analysis and treatment response prediction, which specifically includes the following steps: Step C1: The WGCNA algorithm was used to construct a protein co-expression module. The protein co-expression module included FOXO3, TLR7, and IL6ST. The correlation coefficient r was greater than 0.82, and the significance was P < 0.001. The three proteins had a close synergistic regulatory relationship at the transcriptional level. Step C2: Pathway logic verification; FOXO3, as an upstream regulatory factor, its downregulation inhibits downstream TLR7 transcription. TLR7 further leads to decreased expression of IL6ST, an important signaling molecule for B cell activation. IL6ST's loss prevents B cells from differentiating into plasma cells, ultimately causing a 53% decrease in plasma cell activity and leading to humoral immune failure. Step C3: Set the threshold for targeted therapy. When the molecular levels in patients meet the following conditions, they are defined as the group with advantages in B cell targeted therapy. The specific conditions are as follows: ; in, for threshold for targeted therapy; for threshold for targeted therapy; for threshold for targeted therapy; Step C4: Patients who meet the threshold for targeted therapy receive a TLR7 agonist.

[0015] The present invention discloses a viral ARDS prognosis assessment system and method based on protein combined detection, which has the following beneficial effects: First, by combining serum IL6ST / FOXO3 / TLR7 protein expression levels to reflect B cell functional status, a multivariate logistic regression model was constructed to achieve unified risk stratification for patients of all ages, from children to the elderly. Second, by targeting the IL6ST / FOXO3 / TLR7 pathway, patients with B cell failure can be identified based on pathway activity, and targeted therapy thresholds can be defined to directly guide targeted therapy and facilitate immune intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 Flowchart of the present invention.

[0017] Figure 2 Flowchart of the protein detection module in the present invention. DETAILED DESCRIPTION

[0018] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] like Figure 1 and 2 The viral ARDS prognosis assessment system based on protein combined detection shown includes a sample processing module for achieving serum standardization pretreatment; eliminating pretreatment differences between samples and ensuring consistency of downstream testing; The samples in this example include an ARDS group and a control group; the ARDS group includes 20 patients with severe viral pneumonia who meet the Berlin definition and the Pediatric Acute Lung Injury Consensus Conference criteria, including 11 children under 15 years old and 9 adults over 45 years old; the control group includes 27 subjects, including 14 healthy adults and 13 children undergoing elective surgery without inflammation.

[0020] Continuous blood sampling was performed on ARDS patients, including T1, T2 survivors, and T2 non-survivors. T1 was defined as the first 24 hours after the initiation of mechanical ventilation or ECMO. Blood samples were collected within 24 hours of the start of mechanical ventilation or ECMO life support to obtain physiological indicators of patients in the early stages of life support treatment, which served as baseline data for comparison with subsequent changes in the condition. For T2 survivors, blood samples are collected when mechanical ventilation or ECMO is successfully withdrawn for weaning from life support. This is used to reflect the patient's physical condition at the time of withdrawal and to assess treatment efficacy or prognosis. T2 non-survivors are those whose condition worsens and who cannot maintain life support before the termination of extracorporeal support. Blood samples are collected when it is decided to terminate ECMO for analysis of the pathophysiological changes in the terminal stage.

[0021] Protein detection module for iBAQ quantification of IL6ST / FOXO3 / TLR7 based on DIA mass spectrometry; achieving targeted quantification of core proteins and avoiding ELISA platform bias. At the biological level, the progressive decline of B cell function with aging, namely immunosenescence, has been identified as a key factor; studies in Aging Cell (2009) and Blood (2011) showed that B cell diversity in the elderly decreased by 40%-60%, and CD27+ memory B cells decreased significantly, resulting in a baseline immunoglobulin library diversity that was significantly lower than that of children; especially during acute infection, this defect directly weakened the antiviral immune response; in terms of molecular mechanism, Nature Cell Biology (Nat Cell Biol, 2025) pointed out that FOXO3 affects survival by regulating B cell apoptosis resistance and oxidative stress response; Nature Immunol (Nat Immunol, 2022) confirmed that TLR7 acts as a viral ssRNA pattern recognition receptor, driving B cell antiviral activation; Journal of Clinical Immunology (JClin Immunol, 2023) pointed out that IL6ST, as a core component of IL-6 signaling, is an essential protein for plasma cell differentiation and antibody production.

[0022] iBAQ is an intensity-based absolute quantification, where the peptide signal intensity detected by mass spectrometry reflects the expression level of the protein; the ELISA platform is a quantitative detection technology platform based on the principle of antigen-antibody specific binding, combined with enzymatic reaction to amplify the signal.

[0023] DIA mass spectrometry, data-independent acquisition, is a non-discriminatory and non-random proteomic analysis technology that divides the full mass spectrometer scan range into several windows, and then detects and fragments all ions in each window, thereby obtaining information on all ions in the sample without omission or difference, reducing missing values ​​in sample detection, while improving the quantification area and repeatability, and achieving highly stable and accurate proteomic quantitative analysis in large sample cohorts.

[0024] IL6ST is an interleukin-6 signal transduction protein that can promote B cell differentiation and plasma cell antibody production, and participates in the pathological process of autoimmune diseases; FOXO3 is a forkhead box transcription factor O3, which can participate in DNA damage repair and oxidative stress resistance, and delay cell aging; TLR7 is Toll-like receptor 7, which can recognize single-stranded RNA, such as viral genomic RNA or degradation products, and activate the innate immune response.

[0025] The protein detection module uses DIA mass spectrometry targeted quantitative technology. Its workflow includes the following steps: Step A1: The digested peptides are concentrated and dried using a SpeedVac vacuum centrifuge to obtain the peptide sample. SpeedVac is a vacuum centrifugal concentrator, a laboratory device that uses a vacuum environment, centrifugal force, and heating to rapidly evaporate solvents and concentrate samples. The vacuum lowers the melting and boiling points, while the centrifugal force evenly distributes the sample to increase the evaporation area. Heating further accelerates evaporation, achieving efficient concentration. Step A2: The peptide sample is separated using a high-performance liquid chromatography system. In this example, the EASY LC 1200 is used for separation. High-performance liquid chromatography is a technique that uses a mobile phase to carry the sample through a stationary phase and achieves separation by utilizing the differences in the interactions between different components and the stationary phase. The physicochemical properties of the peptide determine its retention behavior on the stationary phase. By optimizing the mobile phase composition, complex peptide mixtures can be eluted sequentially to form separated chromatographic peaks. The steps specifically include: Step A21: Redissolve the peptide sample in solvent A. In this embodiment, solvent A is a 0.1% formic acid aqueous solution. Step A22: Load the sample onto a self-packing capture column, which can adsorb and retain the target peptide and allow impurities to elute with the mobile phase. The inner diameter of the capture column in this embodiment is 100 μm, which can reduce peptide sample loss and improve the coupling efficiency with mass spectrometry. The filler particle size is 3 μm, which can separate more complex peptide mixtures. Step A23: Separation is performed through an analytical column; in this embodiment, the inner diameter of the analytical column is 150 μm and the column length is 15 cm, which can not only achieve high-resolution separation, but also increase the retention time of the sample in the column and improve the separation effect.

[0026] Step A3: The eluted peptides are peptide samples eluted after separation on the chromatographic column; the eluted peptides are ionized at 2 kV to maintain peptide integrity and are then transferred to the Q Exactive HF-X quadrupole-orbitrap mass spectrometer for data-independent acquisition. The mass spectrometry parameters were as follows: the primary scan (MS1) had a mass-to-charge ratio range of 300-1400 and a resolution of 60,000; the secondary scan (MS2) was segmented into 30 DIA windows, with an acquisition time of 20 ms per window and a resolution of 15,000. The collision energy (i.e., the energy of the precursor ion upon collision with the inert gas in the collision cell) was set to 27%, and the default charge state was set to +3 for calculating the theoretical m / z of the fragment ions. Peptides with a +3 charge state generally require higher collision energies for efficient fragmentation. β-actin was used as the internal reference, and data were normalized by log2 transformation to reduce data scatter and facilitate statistical analysis.

[0027] iBAQ quantification of IL6ST / FOXO3 / TLR7 was performed, specifically including the following steps: Step B1: The raw mass spectrometry data were input into the iProteome cloud platform for processing, and the label-free iBAQ absolute quantification method was used to calculate the total protein ratio × 10 3 Normalize the expression levels, and fill missing values ​​with the minimum non-zero value of 1%. Only proteins with missing values ​​<50% were screened for differential analysis. After log2 transformation, the t-test was used to determine the significantly differentially expressed proteins (P < 0.05). Step B2: Functional enrichment analysis was performed using Metascape (GO biological process P < 0.01); Step B3: 2280 proteins were included in the weighted gene co-expression network (WGCNA) analysis. After variance stabilization and log2(x+1) transformation, a soft threshold β = 2 was selected to construct a topological overlap matrix. The dynamic cutting method was used to identify co-expression modules. The core parameters of the co-expression module were set to a minimum module size of 30, that is, each co-expression module contained at least 30 proteins to avoid the increase of randomness in the results due to too small a module. The merging height was 0.25, that is, based on the similarity between modules, modules were merged when the similarity between modules was ≥ 0.25 to reduce redundant modules. Step B4: Perform cell enrichment analysis using the xCell algorithm and calculate the ssGSEA enrichment score based on 489 cell-specific gene sets. ssGSEA can calculate the enrichment score of a specific gene set for a single sample and quantify the expression activity of the gene set in the sample; Step B5: Prognostic evaluation of FOXO3, IL6ST, and TLR7 proteins; specifically including the following steps:

[0028] Step B51: Screen patients with severe ARDS as the AA group and child survivors as the CA1 group, with death status as the endpoint variable, where 1 represents death and 3 represents survival; Step B52: Univariate and multivariate logistic regression are used to generate execution risk scores. The ROC curve is used to evaluate the discriminant efficacy. A multivariate logistic regression model is constructed through binomial generalized linear modeling to achieve cross-age risk stratification. The formula for the execution risk score in the multivariate logistic regression model is: ; in, is the weight coefficient of IL6ST, and its contribution is 38% according to regression analysis; is the weight coefficient of FOXO3, and its contribution is 42% according to regression analysis; is the weight coefficient of TLR7, and its contribution is 31% according to regression analysis; for risk thresholds; for risk thresholds; for risk thresholds; Measure the IL6ST expression value for the patient. Measure the FOXO3 expression value for the patient. TLR7 expression values ​​were measured for patients; 、 and All are DIA mass spectrometry detection results; if the patient's measured expression value is higher than the threshold, the item is set to 0 to avoid negative contribution.

[0029] In this embodiment, the weight coefficient of IL6ST is determined to have a contribution of 38% through regression analysis; the weight coefficient of FOXO3 is determined to have a contribution of 42% through regression analysis; and the weight coefficient of TLR7 is determined to have a contribution of 31% through regression analysis. The risk threshold is 6.0; The risk threshold is 5.5; The risk threshold is 5.2; the formula for executing the risk R score is: ; If R < 1.2, the patient's mortality rate is < 30%; if R ≥ 1.2, the patient's mortality rate is greater than 70%. Standardized sample processing ensures data comparability, accurately quantifies key proteins, and uses multivariate logistic regression models to achieve cross-age risk stratification, translating protein expression into clinical decision-making signals.

[0030] A multivariate logistic regression model that can achieve cross-age risk scores was constructed, which included the following steps: Step D1: Select training data as a training set and preprocess the training data, including cleaning and standardization. In this embodiment, the training set is 66 samples from the internal cohort. Step D2: Select samples from the GSE225349 dataset as the external validation set, and perform data cleaning and standardization on the data in the external validation set; Step D3: Input the pre-processed training data in the training set into the multivariate logistic regression model for model training and perform iterative optimization to minimize the difference between the probability value predicted by the model and the measured value, thereby obtaining the optimal model parameters; Step D4: Perform model evaluation. In this embodiment, the model evaluation index AUC on the training set reached 0.859, indicating that the model has good discrimination ability on the training data and can effectively identify samples of different states. On the external validation set, the model evaluation index AUC is 0.802, indicating that the model has a certain generalization ability on the new data set and can relatively accurately perform cross-age risk stratification prediction.

[0031] Intelligent analysis module for cross-age risk stratification based on multivariate logistic regression model.

[0032] This embodiment provides the application of patient A and patient B in a multivariate logistic regression model; patient A is an adult, and the actual measured value of IL6ST of patient A by DIA mass spectrometry is 4.8, the actual measured value of FOXO3 is 4.5, and the actual measured value of TLR7 is 4.3. After calculating the risk formula using the multivariate logistic regression model, the risk RiskScore of patient A is 1.52. The prognostic assessment system determines that the risk is high, and patient A has no vital signs. The prognostic assessment result is consistent with the clinical outcome; patient B is a child, and the actual measured value of IL6ST of patient B by DIA mass spectrometry is 5.9, the actual measured value of FOXO3 is 5.8, and the actual measured value of TLR7 is 5.5. After calculating the risk formula using the multivariate logistic regression model, the risk RiskScore of patient B is 0.21. The prognostic assessment system determines that the risk is low, and patient B still maintains a good vital sign status. The post-evaluation result is consistent with the clinical outcome.

[0033] Statistical analysis was performed using R4.3.0 and GraphPad Prism 9.0. Appropriate statistical methods were selected according to the data distribution characteristics and variable types to ensure the reliability of the results.

[0034] After the Shapiro-Wilk normality test, continuous variables were tested with the t-test for normally distributed data. This method infers intergroup significance based on mean differences and requires that the data meet normality and homogeneity of variance. For non-normally distributed data, the Wilcoxon test was used. This method is based on rank comparison rather than raw data value comparison and has no strict requirements on distribution. For categorical variables, if the sample size is large, the χ² test is used to determine the significance by calculating the difference between the actual frequency and the theoretical frequency. If the sample size is small, the Fisher exact test is used to directly calculate the probability value when the total amount is small, thus avoiding the error caused by the χ² test. In this embodiment, the theoretical frequency of the large sample is ≥5, and the theoretical frequency of the small sample is <5.

[0035] High-dimensional data were corrected using Benjamini-Hochberg, and the significance threshold was adjusted by controlling the false discovery rate. In this embodiment, P<0.05 was used as the significance threshold, which can retain the verification efficiency while reducing false positives.

[0036] The algorithm code for implementing cross-age risk stratification based on the multivariate logistic regression model is as follows: def risk_stratification(IL6ST, FOXO3, TLR7): score = 0.38*max(0, 6.0 - IL6ST) + 0.42*max(0, 5.5 - FOXO3) + 0.31*max(0, 5.2 - TLR7) if score < 1.2: return "low risk (mortality rate < 30%)" else: return "High risk (mortality rate > 70%)".

[0037] The intelligent analysis module, including the immune aging pathway integration unit, is used to analyze the mechanism and predict the treatment response, addressing the cognitive bias of the pathological mechanism. Specifically, it includes the following steps: Step C1: The WGCNA algorithm was used to construct a protein co-expression module. The protein co-expression module included FOXO3, TLR7, and IL6ST. The correlation coefficient r was greater than 0.82, and the significance was P < 0.001. The three proteins had a close synergistic regulatory relationship at the transcriptional level. Step C2: Pathway logic verification; FOXO3, as an upstream regulatory factor, its downregulation inhibits downstream TLR7 transcription. TLR7 further leads to decreased expression of IL6ST, an important signaling molecule for B cell activation. IL6ST's loss prevents B cells from differentiating into plasma cells, ultimately causing a 53% decrease in plasma cell activity and leading to humoral immune failure. Step C3: Set the threshold for targeted therapy. When the molecular levels in patients meet the following conditions, they are defined as the group with advantages in B cell targeted therapy. The specific conditions are as follows:

[0038] in, for threshold for targeted therapy; for threshold for targeted therapy; for threshold for targeted therapy; Step C4: Patients who meet the threshold for targeted therapy receive a TLR7 agonist.

[0039] This example can effectively activate the B cell pathway through TLR7 agonists. When TLR7 agonists are used to treat people who meet the threshold, the efficacy is increased by 3.2 times. It solves the existing problem of focusing only on the direct effects of inflammatory factors, assuming that "intensity of inflammation is positively correlated with disease mortality", but cannot explain the clinical paradox of "high inflammation and low mortality in children", that is, children have high levels of inflammatory factors but a mortality rate significantly lower than that of adults; it quantifies the age-dependent functional heterogeneity of B cells. In pediatric patients, the level of B cell activation can be increased by 2.8 times compared with that of adult patients. Even in the presence of a high inflammatory state, pathogens can still be eliminated through a strong humoral immune response, thereby reducing the risk of death; in adult patients, due to immune aging, the function of the FOXO3-TLR7-IL6ST pathway declines, leading to B cell exhaustion and a 53% decrease in plasma cell activity. Even if the level of inflammatory factors is low, the prognosis may be poor due to insufficient immune response.

[0040] The prognostic assessment method for viral ARDS based on combined protein detection specifically includes the following steps: Step S1: receiving a sample; i.e., receiving a patient serum sample with a volume of 50-100 μL, and introducing the serum into the sample processing module of the ARDS prognosis assessment system via an automatic sample injector for serum standardization; collecting 3 ml of venous blood into an EDTA anticoagulant tube, and incubating at 20° C. for 20 minutes; the EDTA anticoagulant tube in this embodiment is BD#367861; Step S2: The serum sample is placed in a constant temperature metal bath; that is, the sample is heated in an aluminum heat block for three minutes to achieve heat denaturation and inactivation of the protein. In this embodiment, the aluminum heat block is controlled at a temperature of 94°C to 96°C, and the serum sample is inactivated at 95°C for 3 minutes. Specifically, 2 μL of plasma is mixed with 98 μL of 50 mM ABC buffer and heat inactivated at 95°C for 3 minutes. Step S3: Entering the digestion reaction chamber; the digestion reaction chamber is a constant temperature oscillating chamber at 37°C, and pancreatic enzyme digestion is performed inside the digestion reaction chamber; wherein the oscillation frequency is 120 rpm, the mass ratio of enzyme to protein is 1:25, and the reaction time is 12 hours; that is, pancreatic enzyme is added at a mass ratio of enzyme to protein of 1:25, that is, 25 units of protein for every 1 unit of pancreatic enzyme, and the digestion is carried out at 37°C overnight; Step S4: Peptide separation by liquid chromatography on an EASY nLC 1200 system using a ReproSil-Pur C18-AQ column at a flow rate of 300 nL / min and a gradient of 5-35% acetonitrile / 120 min. Step S5: performing mass spectrometry detection; performing DIA scanning based on a QE HF-X mass spectrometer; wherein the scanning range is m / z 300-1400; the MS1 resolution is 60,000, the MS2 resolution is 15,000; and the collision energy is 27%; Step S6: iBAQ quantification of IL6ST / FOXO3 / TLR7 based on the iProteome cloud platform; Step S7: Calculate the risk score based on the multivariate logistic regression model in the intelligent analysis module and output the risk level; determine the size of the risk. If the risk score is less than 1.2, the risk level is output as low risk, and the indicator light is green; if the risk score is greater than or equal to 1.2, the risk level is output as high risk, and the indicator light is red.

[0041] The present invention has the following beneficial effects: First, a universal prognostic assessment system for viral ARDS that is applicable across all age groups. This system combines serum IL6ST / FOXO3 / TLR7 protein expression levels to reflect B cell function, constructing a multivariate logistic regression model to achieve unified risk stratification for patients of all ages, from children to the elderly. Second, by targeting immune aging-specific pathways and mediating B cell apoptosis through FOXO3, the model inhibits TLR7 expression. This cascade of IL6ST dysfunction is a key molecular mechanism driving age-related heterogeneity in prognosis. The model evaluation metric, AUC, increased to 0.859, significantly outperforming the existing technology's 68.9%, creating a golden window for clinical intervention. 3. Establish a treatment response prediction system; existing technologies only provide risk probabilities, while the biological thresholds defined in this invention directly guide targeted therapy. Patients who meet the high-risk threshold have a 2.1-fold increase in survival rate after receiving TLR7 agonists.

[0042] The above embodiments are not limitations of the present invention, and the present invention is not limited to the above examples. Any changes, modifications, additions or substitutions made by technicians in this technical field within the scope of the technical solution of the present invention also fall within the scope of protection of the present invention.

Claims

1. A viral ARDS prognosis assessment system based on protein combined detection, characterized by: It includes a sample processing module for pre-treatment of serum standardization; a protein detection module for iBAQ quantification of IL6ST / FOXO3 / TLR7 based on DIA mass spectrometry to achieve targeted quantification of core proteins; Intelligent analysis module for cross-age risk stratification based on multivariate logistic regression model.

2. The viral ARDS prognosis assessment system based on protein combined detection according to claim 1, characterized in that: The intelligent analysis module includes an immune aging pathway integration unit, which is used to perform mechanism analysis and treatment response prediction to solve the problem of cognitive bias in pathological mechanisms.

3. A method for evaluating the prognosis of viral ARDS based on combined protein detection, characterized in that: The evaluation method is applied to the evaluation system according to any one of claims 1 to 2, and the evaluation method specifically comprises the following steps: Step S1: receiving a patient's serum sample and introducing it into the sample processing module of the ARDS prognostic evaluation system through an automatic sample injector for serum standardization; Step S2: The serum sample enters the constant temperature metal bath in the protein detection module to achieve thermal denaturation and inactivation of the protein; Step S3: Entering the digestion reaction chamber; the digestion reaction chamber is a constant temperature oscillation chamber at 37°C, and trypsin digestion is performed inside the digestion reaction chamber; wherein the oscillation frequency is 120 rpm, the mass ratio of enzyme to protein is 1:25, and the reaction time is 12 hours; Step S4: Peptide separation by liquid chromatography based on the EASY nLC 1200 system; Step S5: performing DIA mass spectrometry detection based on QE HF-X; Step S6: iBAQ quantification of IL6ST / FOXO3 / TLR7 based on the iProteome cloud platform; Step S7: Calculate the cross-age risk score based on the multivariate logistic regression model in the intelligent analysis module, and make a risk judgment to output the risk level; the formula for executing the risk score in the multivariate logistic regression model is: ； in, is the weight coefficient of IL6ST; is the weight coefficient of FOXO3; is the weight coefficient of TLR7; for risk thresholds; for risk thresholds; for risk thresholds; Measure the IL6ST expression value for the patient. Measure the FOXO3 expression value for the patient. The TLR7 expression value was measured for the patients.

4. The method for prognosis assessment of viral ARDS based on protein combined detection according to claim 3, characterized in that: The chromatographic column for liquid chromatography peptide separation in step S4 is ReproSil-Pur C18-AQ; the flow rate is 300 nL / min, and the gradient is 5-35% acetonitrile / 120 min.

5. The method for prognosis assessment of viral ARDS based on protein combined detection according to claim 3, characterized in that: In step S5, the scanning range of the DIA mass spectrometry detection based on QE HF-X is m / z 300-1400; the MS1 resolution is 60,000, the MS2 resolution is 15,000; and the collision energy is 27%.

6. The method for prognosis assessment of viral ARDS based on protein combined detection according to claim 5, characterized in that: The DIA mass spectrometry detection is performed in step S5, which specifically includes the following steps: Step A1: The digested peptides were concentrated and dried by SpeedVac vacuum centrifugation to obtain peptide samples; Step A2: The peptide samples are separated by a high performance liquid chromatography system; Step A3: The eluted peptides are peptide samples eluted after separation on the chromatographic column; the eluted peptides are ionized at a voltage of 2 kV to maintain the integrity of the peptides; and are then transferred to the Q Exactive HF-X quadrupole-orbitrap mass spectrometer for data-independent acquisition.

7. The method for prognosis assessment of viral ARDS based on protein combined detection according to claim 6, characterized in that: In step A2, the peptide sample is separated by a high performance liquid chromatography system, which specifically includes the following steps: Step A21: re-dissolving the peptide sample in a solvent; Step A22: Load the sample onto a self-packed trapping column to adsorb and retain the target peptide, allowing impurities to elute with the mobile phase; Step A23: Separation is performed through an analytical column.

8. The method for prognosis assessment of viral ARDS based on protein combined detection according to claim 3, characterized in that: The step S6 performs iBAQ quantification of IL6ST / FOXO3 / TLR7, specifically comprising the following steps: Step B1: The raw mass spectrometry data were input into the iProteome cloud platform for processing, and the label-free iBAQ absolute quantification method was used to calculate the total protein ratio × 10 3 Normalize the expression levels, and fill missing values ​​with the minimum non-zero value of 1%. Only proteins with missing values ​​< 50% were screened for differential analysis, and the significantly differentially expressed proteins were determined by t-test after log2 transformation. Step B2: Complete functional enrichment analysis using Metascape; Step B3: Through weighted gene co-expression network WGCNA analysis, after variance stabilization and log2(x+1) transformation, a soft threshold β=2 was selected to construct a topological overlap matrix; and a dynamic cutting method was used to identify co-expression modules. The minimum number of modules in the core parameters of the co-expression module was 30. When the similarity between modules was ≥0.25, they were merged to reduce redundant modules; Step B4: Perform cell enrichment analysis using the xCell algorithm and calculate the ssGSEA enrichment score based on the cell signature gene set. ssGSEA can calculate the enrichment score of a specific gene set for a single sample and quantify the expression activity of the gene set in the sample; Step B5: Prognostic evaluation of FOXO3, IL6ST, and TLR7 proteins.

9. The method for prognosis assessment of viral ARDS based on protein combined detection according to claim 3, characterized in that: In step S7, the cross-age risk scoring is performed based on the multivariate logistic regression model in the intelligent analysis module, which can realize the method of constructing the multivariate logistic regression model for cross-age risk scoring, and specifically includes the following steps: Step D1: Select training data as the training set and preprocess the training data. The preprocessing includes cleaning and standardization. Step D2: Select samples from the GSE225349 dataset as the external validation set, and perform data cleaning and standardization on the data in the external validation set; Step D3: Input the pre-processed training data in the training set into the multivariate logistic regression model for model training and perform iterative optimization to minimize the difference between the probability value predicted by the model and the measured value, thereby obtaining the optimal model parameters; Step D4: Calculate the model evaluation indicator AUC.

10. The method for prognosis assessment of viral ARDS based on protein combined detection according to claim 9, characterized in that: The immunosenescence pathway integration unit performs mechanism analysis and therapeutic response prediction, specifically including the following steps: Step C1: The WGCNA algorithm was used to construct a protein co-expression module. The protein co-expression module included FOXO3, TLR7, and IL6ST. The correlation coefficient r was greater than 0.82, and the significance was P < 0.

001. The three proteins had a close synergistic regulatory relationship at the transcriptional level. Step C2: Pathway logic verification; FOXO3, as an upstream regulatory factor, its downregulation inhibits downstream TLR7 transcription. TLR7 further leads to decreased expression of IL6ST, an important signaling molecule for B cell activation. IL6ST's loss prevents B cells from differentiating into plasma cells, ultimately causing a 53% decrease in plasma cell activity and leading to humoral immune failure. Step C3: Set the threshold for targeted therapy. When the molecular levels in patients meet the following conditions, they are defined as the group with advantages in B cell targeted therapy. The specific conditions are as follows: ； in, for threshold for targeted therapy; for threshold for targeted therapy; for threshold for targeted therapy; Step C4: Patients who meet the threshold for targeted therapy receive a TLR7 agonist.

Images