A bronchial asthma prediction method and system

Abstract

The present invention discloses a bronchial asthma prediction method and system, which relates to the technical field of asthma prediction, including defining an experimental group and a control group and extracting RNA from the experimental group and the control group for cryopreservation; performing quantitative PCR analysis on the cryopreserved RNA and extracting target cDNA; collecting clinical data of the experimental group and calculating the corresponding biomarker concentration according to the target cDNA; constructing a bronchial asthma prediction model, and outputting the predicted risk probability value of each patient suffering from asthma. By accurately extracting and preserving RNA samples from the experimental group and the control group, combining quantitative PCR analysis to obtain significantly changed gene expression, and further combining with biomarker concentrations, the dynamic monitoring of asthma-related inflammatory reactions is successfully achieved. By constructing a bronchial asthma prediction model, the invention can accurately assess the patient's asthma risk.

Description

Technical Field

[0001] The present invention relates to the technical field of asthma prediction, and in particular to a bronchial asthma prediction method and system. Background Art

[0002] With the continuous advancement of biomedical technology, the treatment of bronchial asthma has made significant progress. At present, the routine diagnosis of bronchial asthma mainly relies on clinical symptoms, pulmonary function tests, and bronchial provocation tests. However, traditional diagnostic methods often cannot fully reveal the molecular biological data of asthma, especially the changes in gene expression involving inflammatory pathways and immune responses. This limitation makes the diagnostic results easily affected by the complexity of clinical manifestations. In recent years, with the application of molecular diagnostic technology, researchers have begun to pay attention to diagnostic methods based on gene expression, such as qPCR analysis and biomarker detection, aiming to reveal the activity of patients' airway inflammatory response through changes in gene expression, and provide new ideas for early detection and personalized treatment of asthma.

[0003] In the existing technology, although qPCR analysis and biomarker detection have been applied to asthma prediction in some studies, these technologies still face many shortcomings in actual clinical applications. The application of qPCR analysis is often limited to specific genes, and it is difficult to comprehensively cover the various inflammatory markers related to asthma, resulting in limited prediction accuracy. The existing detection technology lacks long-term monitoring of the dynamic changes of gene expression, and it is difficult to timely track the changes in the inflammatory state of patients during treatment, thereby delaying the implementation of personalized treatment. In terms of model construction, the existing asthma prediction model is relatively simple, and it is usually difficult to integrate multi-dimensional patient genetic data, clinical data and biomarker information, resulting in limited prediction ability for asthma. How to improve the comprehensiveness, accuracy and dynamics of genetic testing in the prediction process of bronchial asthma has become an urgent problem to be solved. Summary of the invention

[0004] In view of the problems existing in the above-mentioned existing bronchial asthma prediction methods and systems, the present invention is proposed.

[0005] Therefore, the problem to be solved by the present invention is that the existing bronchial asthma prediction technology has problems such as low accuracy and lack of long-term dynamic monitoring;

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a bronchial asthma prediction method, comprising:

[0007] The experimental and control groups were defined and RNA was extracted from the experimental and control groups for cryopreservation;

[0008] The cryopreserved RNA was subjected to quantitative PCR analysis and target cDNA was extracted;

[0009] Collect clinical data of the experimental group and calculate the corresponding biomarker concentration based on the target cDNA;

[0010] Construct a bronchial asthma prediction model and output the predicted risk probability value of asthma for each patient;

[0011] Store and manage all data in a relational database;

[0012] The method of constructing a bronchial asthma prediction model and outputting a predicted risk probability value of each patient suffering from asthma comprises: integrating the clinical data collected in the experimental group and the extracted target cDNA and biomarker concentration into a data set;

[0013] Divide the dataset into training and testing sets;

[0014] Use CatBoost as the classifier model to build a bronchial asthma prediction model and set hyperparameters;

[0015] The setting of hyperparameters includes setting the number of iterations, the learning rate, and the depth of the decision tree;

[0016] The training set was input into the bronchial asthma prediction model. The training process included multiple iterations. Multi-classification cross entropy loss was selected as the loss function. The loss function was optimized by gradient descent in each iteration.

[0017] After the bronchial asthma prediction model training is completed, the performance of the model is evaluated using the test set;

[0018] Use cross-validation to verify the impact of different hyperparameter combinations on model performance and select the optimal hyperparameters;

[0019] The bronchial asthma prediction model is used to calculate all the patient's data through each decision tree to obtain the initial predicted probability of asthma risk M. i ;

[0020] Based on the target cDNA extracted from the data set, the gene expression part is constructed to capture the risk signals of patients at the genetic level. The formula is:

[0021]

[0022] in, represents the contribution of gene expression to the asthma risk of the i-th patient, represents the relative expression level of NLRP3 in the i-th patient, represents the relative expression of IL-1β in the i-th patient, represents the relative expression level of Caspase-1 in the i-th patient;

[0023] Using the biomarker concentrations in the dataset, the biomarker concentration part was constructed using the formula:

[0024]

[0025] in, represents the impact of the biomarker concentration on the risk prediction of the i-th patient, represents the biomarker concentration of TNF-α of the i-th patient, represents the biomarker concentration of IL-6 of the i-th patient, represents the biomarker concentration of IL-1β of the i-th patient;

[0026] The clinical data part is constructed using the patient's clinical data. The formula is:

[0027]

[0028] in, represents the contribution of the clinical data to the asthma risk of the i-th patient, represents the allergy history of the i-th patient, represents the family asthma history of the i-th patient, represents the medical history record of the i-th patient;

[0029] Construct the final risk probability formula to predict the patient's asthma risk probability. The formula is:

[0030]

[0031] in, represents the final predicted probability value of asthma risk for the i-th patient, Represents the Sigmoid function;

[0032] According to the output predicted risk probability value of each patient suffering from asthma, a low risk probability threshold T is set. A With high risk probability threshold T B , classifying the predicted risk probability values ​​based on the risk probability threshold;

[0033] like , indicating that the patient is a low-risk asthma patient;

[0034] like , indicating that the patient is a medium-risk asthma patient;

[0035] like , indicating that the patient is a high-risk asthma patient;

[0036] The classification of low-risk asthma patients, medium-risk asthma patients, and high-risk asthma patients was ranked from low to high;

[0037] Each patient is sampled regularly and integrated into a new data set to be input into the bronchial asthma prediction model to output a new predicted risk probability value.

[0038] As a preferred embodiment of the bronchial asthma prediction method of the present invention, wherein: the steps of defining the experimental group and the control group and extracting RNA from the experimental group and the control group for cryopreservation respectively include: defining the experimental group and the control group respectively according to the experimental requirements;

[0039] The experimental group refers to patients who were judged to be suspected of asthma;

[0040] The control group refers to healthy individuals;

[0041] The nasal cavity, pharynx, and airway of the members of the experimental and control groups were locally anesthetized with 1% lidocaine, and a fiberoptic bronchoscope was slowly inserted into the airway;

[0042] Through a fiberoptic bronchoscope, a disposable cell brush is inserted into the bronchus to collect epithelial cell samples on the surface of the bronchial epithelium;

[0043] The collected epithelial cell samples are placed in a test tube containing a preservation solution, and RNA in the epithelial cell samples is extracted using an RNA extraction kit;

[0044] The purity and concentration of RNA were tested using a spectrophotometer;

[0045] The detection of RNA purity and concentration includes using a spectrophotometer to evaluate RNA purity through an absorbance ratio, and according to the purity evaluation standard, setting the normal range of the absorbance ratio to be between 1.8 and 2.1. If the absorbance ratio of the RNA is not within the normal range, it means that the RNA purity does not meet the standard, and then using an ethanol precipitation method to re-purify the RNA. If the absorbance ratio of the RNA is within the normal range, it means that the RNA purity meets the standard, and then performing a concentration test;

[0046] According to the experimental requirements, the normal concentration range of RNA is set. If the concentration of RNA after the purity test is not within the normal concentration range of RNA, it means that the RNA concentration does not meet the standard, then the RNA sample is concentrated and the RNA is re-extracted. If the concentration of RNA after the purity test is within the normal concentration range of RNA, it means that the RNA concentration meets the standard, then low-temperature storage treatment is performed.

[0047] As a preferred embodiment of the bronchial asthma prediction method of the present invention, wherein: the quantitative PCR analysis of the cryopreserved RNA and the extraction of the target cDNA comprises using a reverse transcription kit to perform a reverse transcription reaction from RNA to cDNA, and after the reverse transcription reaction is completed, a cDNA for quantitative PCR analysis is obtained;

[0048] A single reaction system suitable for quantitative PCR analysis was configured, and SYBR Green method was used as the detection method;

[0049] The configuration of the single reaction system includes configuring SYBR Green Master Mix premix reagent, template cDNA, forward primer and reverse primer;

[0050] The cDNA targeted by the forward primer and the reverse primer in a single reaction system is designed and called the target cDNA;

[0051] The target cDNAs include NLRP3, IL-1β, and Caspase-1;

[0052] After the single reaction system is configured, the target cDNA is added as a template into the single reaction system to obtain a configured qPCR reaction mixture, and the qPCR reaction mixture is placed on a qPCR reaction plate and sealed with a transparent optical sealing film;

[0053] Place the sealed reaction plate into a real-time fluorescence quantitative qPCR instrument, set the reaction program and perform qPCR amplification reaction;

[0054] The reaction program includes initial deformation, cyclic reaction and dissolution curve;

[0055] A predetermined threshold is set by the intensity of the fluorescence signal, the number of cycles that first exceed the predetermined threshold in each cycle of the experimental group and the control group is monitored, and the number of cycles that first exceed the predetermined threshold is taken as the CT value;

[0056] According to the obtained CT value, the CT value difference between the target cDNA and the internal reference cDNA in the experimental group and the control group was calculated respectively;

[0057] According to the obtained CT value difference between the target cDNA and the internal reference cDNA, the difference in CT value between the experimental group and the control group was calculated;

[0058] The relative expression levels of the target cDNA in the experimental group and the control group were calculated.

[0059] As a preferred embodiment of the bronchial asthma prediction method of the present invention, wherein: the collecting clinical data of the experimental group and calculating the corresponding biomarker concentration according to the target cDNA comprises: collecting the clinical data of the experimental group from the electronic medical record system of the hospital;

[0060] The clinical data included allergy history, family asthma history, and medical history records;

[0061] If you have a history of allergies, please record your allergy history. g,i is 1, otherwise it is 0;

[0062] If there is a family history of asthma, please record the family history of asthma. j,i is 1, otherwise it is 0;

[0063] If you have a history of asthma, please record the medical history. z,i is 1, otherwise it is 0;

[0064] Collect serum samples from the experimental group and process the samples. Dilute the processed serum samples in proportion according to experimental needs and add them to a 96-well plate.

[0065] Based on the target cDNA, the absorbance value of the biomarker in the serum sample was detected by enzyme-linked immunosorbent assay, and the concentration of each biomarker was calculated by comparing the absorbance with the standard curve of standards with known concentrations;

[0066] The biomarker concentrations include the concentrations of TNF-α, IL-6, and IL-1β in serum.

[0067] As a preferred solution of the bronchial asthma prediction method described in the present invention, the storing and managing of all data in a relational database includes selecting a relational database to store and manage the patient's data set and its analysis results, designing a database table structure to store different types of data, setting regular backup tasks, backing up all data in the database, performing authority management on database users, and encrypting and storing sensitive data.

[0068] Another object of the present invention is to provide a bronchial asthma prediction system, which comprises:

[0069] RNA extraction module, used to define the experimental group and the control group and extract RNA from the experimental group and the control group for cryopreservation;

[0070] cDNA selection module, used to perform quantitative PCR analysis on cryopreserved RNA and extract target cDNA;

[0071] The concentration calculation module is used to collect clinical data of the experimental group and calculate the corresponding biomarker concentration based on the target cDNA;

[0072] A model building module is used to build a bronchial asthma prediction model and output a predicted risk probability value of asthma for each patient;

[0073] The data storage module is used to store and manage all data in a relational database.

[0074] A computer device comprises: a memory and a processor; the memory stores a computer program, and the processor implements the steps of the bronchial asthma prediction method as described above when executing the computer program.

[0075] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the bronchial asthma prediction method as described above.

[0076] The beneficial effects of the present invention are: by accurately extracting and preserving RNA samples from the experimental group and the control group, combining quantitative PCR analysis to obtain significantly changed gene expression, and further combining it with biomarker concentrations, dynamic monitoring of asthma-related inflammatory responses is successfully achieved. By constructing a bronchial asthma prediction model, the invention can accurately assess the patient's asthma risk. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0078] Figure 1 Schematic diagram of the flow chart of the bronchial asthma prediction method.

[0079] Figure 2 Schematic diagram of the structure for predicting the risk probability value of a patient.

[0080] Figure 3 Schematic diagram of the structure of the bronchial asthma prediction system. DETAILED DESCRIPTION

[0081] In order to make the above-mentioned objects, data and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0082] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0083] Secondly, the term "one embodiment" or "embodiment" as used herein refers to specific data, structures or characteristics that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments. Example

[0084] Reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and which provides a bronchial asthma prediction method, the bronchial asthma prediction method comprising:

[0085] S1, define the experimental group and the control group and extract the RNA of the experimental group and the control group for cryopreservation;

[0086] Specifically, defining the experimental group and the control group and extracting RNA from the experimental group and the control group for cryopreservation respectively includes: defining the experimental group and the control group respectively according to the experimental requirements;

[0087] The experimental group refers to patients who were judged to be suspected of asthma;

[0088] The control group refers to healthy individuals;

[0089] The nasal cavity, pharynx, and airway of the members of the experimental and control groups were locally anesthetized with 1% lidocaine, and a fiberoptic bronchoscope was slowly inserted into the airway;

[0090] Through a fiberoptic bronchoscope, a disposable cell brush is inserted into the bronchus to collect epithelial cell samples on the surface of the bronchial epithelium;

[0091] The collected epithelial cell samples are placed in a test tube containing a preservation solution, and RNA in the epithelial cell samples is extracted using an RNA extraction kit;

[0092] The extracting RNA from the epithelial cell sample comprises centrifuging the collected epithelial cell sample, removing the preservation solution, adding the reagent in the RNA extraction kit, and fully mixing, standing for 5 minutes, adding chloroform, fully mixing and centrifuging, extracting the supernatant, adding an equal volume of isopropanol to precipitate RNA, washing the RNA precipitate after centrifugation, and finally washing with anhydrous ethanol, and then dissolving the RNA in RNase-free water;

[0093] The purity and concentration of RNA were tested using a spectrophotometer;

[0094] The detection of RNA purity and concentration includes using a spectrophotometer to evaluate RNA purity through an absorbance ratio, and according to the purity evaluation standard, setting the normal range of the absorbance ratio to be between 1.8 and 2.1. If the absorbance ratio of the RNA is not within the normal range, it means that the RNA purity does not meet the standard, and then using an ethanol precipitation method to re-purify the RNA. If the absorbance ratio of the RNA is within the normal range, it means that the RNA purity meets the standard, and then performing a concentration test;

[0095] The ethanol precipitation method is a commonly used method for purifying and concentrating nucleic acids (RNA or DNA), which uses ethanol to precipitate nucleic acids from a solution, thereby removing impurities and increasing the concentration of nucleic acids;

[0096] According to the experimental requirements, the normal concentration range of RNA is set. If the concentration of RNA after the purity test is not within the normal concentration range of RNA, it means that the RNA concentration does not meet the standard, then the RNA sample is concentrated and the RNA is re-extracted. If the concentration of RNA after the purity test is within the normal concentration range of RNA, it means that the RNA concentration meets the standard, then low-temperature storage treatment is performed.

[0097] Fiberbronchoscope technology was used to collect samples from the bronchial epithelium of the experimental group and the control group to ensure the consistency of the source of the cell samples. Through multiple steps of RNA extraction and detection, impurities were effectively removed, ensuring that the purity and concentration of RNA were within the set standard range. This method used a spectrophotometer to detect the purity and concentration of RNA, and used ethanol precipitation for further purification, which improved the quality of RNA samples and ensured the accuracy and reliability of subsequent analysis. The low-temperature preservation step of RNA allowed the samples to be stored for a long time for subsequent gene expression analysis, thereby improving the repeatability of the experiment and the stability of the data.

[0098] S2, the cryopreserved RNA was subjected to quantitative PCR analysis and target cDNA was extracted;

[0099] Specifically, performing quantitative PCR analysis on the cryopreserved RNA and extracting target cDNA includes using a reverse transcription kit to perform a reverse transcription reaction from RNA to cDNA, and obtaining cDNA for quantitative PCR analysis after the reverse transcription reaction is completed;

[0100] The reverse transcription reaction includes incubating the RNA sample at 25°C for 5 minutes (primer binding to RNA), incubating at 55°C for 30-60 minutes (reverse transcription), and heating at 70°C for 10 minutes (termination reaction);

[0101] A single reaction system suitable for quantitative PCR analysis was configured, and SYBR Green method was used as the detection method;

[0102] The configuration of the single reaction system includes configuring SYBR Green Master Mix premix reagent, template cDNA, forward primer and reverse primer;

[0103] The SYBR Green Master Mix premix reagent includes SYBR Green dye, dNTP, Taq DNA polymerase and reaction buffer;

[0104] The cDNA targeted by the forward primer and the reverse primer in a single reaction system is designed and called the target cDNA;

[0105] The target cDNAs include NLRP3, IL-1β, and Caspase-1;

[0106] The SYBR Green method combined with primer design can effectively detect the expression of target cDNA. The SYBR Green dye emits fluorescence when it binds to double-stranded DNA, and the progress of amplification can be monitored in real time.

[0107] After the single reaction system is configured, the target cDNA is added as a template into the single reaction system to obtain a configured qPCR reaction mixture, and the qPCR reaction mixture is placed on a qPCR reaction plate and sealed with a transparent optical sealing film;

[0108] Place the sealed reaction plate into a real-time fluorescence quantitative qPCR instrument, set the reaction program and perform qPCR amplification reaction;

[0109] The reaction program includes initial deformation (used to melt double-stranded DNA into single strands), cyclic reaction (each cycle includes DNA melting, primer binding and DNA amplification) and melting curve (used to verify the specificity of the amplified product, and detect whether there is non-specific amplified product through the change of fluorescence signal at different temperatures);

[0110] A predetermined threshold is set by the intensity of the fluorescence signal, the number of cycles that first exceed the predetermined threshold in each cycle of the experimental group and the control group is monitored, and the number of cycles that first exceed the predetermined threshold is taken as the CT value;

[0111] The CT value is a value automatically generated by the qPCR instrument based on the relationship between the fluorescence signal and the number of cycles. In each qPCR amplification reaction cycle, as the amount of target cDNA increases, the fluorescence signal gradually increases until it exceeds a predetermined threshold for the first time. The number of cycles when this threshold is first reached is the CT value.

[0112] According to the obtained CT value, the CT value difference between the target cDNA and the internal reference cDNA in the experimental group and the control group was calculated respectively, and the expression was:

[0113] ,

[0114] in, Indicates the CT value difference between the target cDNA and the internal reference cDNA, CT b Indicates the CT value of the target cDNA, CT c It represents the CT value of the internal reference cDNA, which is used for standardization. Its expression level remains relatively stable in different samples and is often used to calibrate the expression of the target cDNA.

[0115] According to the CT value difference between the target cDNA and the internal reference cDNA, the difference in CT value between the experimental group and the control group was calculated. The expression is:

[0116] ,

[0117] in, It represents the difference in CT value between the experimental group and the control group. represents the CT value difference between the target cDNA and the internal reference cDNA of the xth patient in the experimental group, represents the CT value difference between the target cDNA and the internal reference cDNA of the yth healthy individual in the control group;

[0118] The relative expression of the target cDNA in the experimental group and the control group was calculated using the formula:

[0119] ,

[0120] Among them, F x,y It indicates the relative expression level of target cDNA in the experimental group and the control group.

[0121] By reverse transcribing RNA to generate target cDNA and using the SYBR Green method combined with primer design, the expression of the target gene can be monitored in real time with high efficiency and accuracy. During the qPCR amplification process, the fluorescent signal is proportional to the expression level of the target gene. Using the CT value as a quantitative standard, the gene expression levels of the experimental group and the control group can be compared, thereby accurately evaluating the relative expression levels of the gene. This method greatly improves the detection efficiency of genes related to diseases such as asthma and enhances the accuracy and specificity of the experiment.

[0122] S3, collecting clinical data of the experimental group and calculating the corresponding biomarker concentration based on the target cDNA;

[0123] Specifically, collecting clinical data of the experimental group and calculating the corresponding biomarker concentration based on the target cDNA includes: collecting clinical data of the experimental group from the electronic medical record system of the hospital;

[0124] The clinical data included allergy history, family asthma history, and medical history records;

[0125] If you have a history of allergies, please record your allergy history. g,i is 1, otherwise it is 0;

[0126] If there is a family history of asthma, please record the family history of asthma. j,i is 1, otherwise it is 0;

[0127] If you have a history of asthma, please record the medical history. z,i is 1, otherwise it is 0;

[0128] Collect serum samples from the experimental group and process the samples. Dilute the processed serum samples in proportion according to experimental needs and add them to a 96-well plate.

[0129] The sample processing includes placing the collected blood sample in a blood collection tube without anticoagulant for 20-30 minutes to promote natural coagulation of the blood, centrifuging at 3000 rpm for 10 minutes at 4°C to separate the serum, and the obtained serum should be quickly transferred to a sterile tube, marked with the numbers of the experimental group and the control group, and stored in an environment of -80°C to ensure that the sample will not degrade before testing;

[0130] Based on the target cDNA, the absorbance value of the biomarker in the serum sample was detected by enzyme-linked immunosorbent assay, and the concentration of each biomarker was calculated by comparing the absorbance with the standard curve of standards with known concentrations;

[0131] The enzyme-linked immunosorbent assay refers to a process of quantitatively analyzing the concentration of a marker through specific antigen-antibody binding and enzymatic reaction to amplify the signal. During the detection process, a specific antibody is used to bind to the corresponding marker, and the enzyme-labeled antibody produces a color change through a substrate reaction. The intensity of the color is proportional to the concentration of the marker in the sample;

[0132] The standard curve of the known concentration standard refers to a tool used to determine the concentration of the target substance (such as a biomarker) in an unknown sample. The standard curve is generated by a series of known concentrations of standards (i.e., solutions containing known amounts of antigens or antibodies to be tested), and describes the relationship between absorbance (OD value) and known concentrations;

[0133] The biomarker concentrations include the concentrations of TNF-α, IL-6, and IL-1β in serum.

[0134] This method ensures accurate assessment of the concentration of asthma-related biomarkers (such as TNF-α, IL-6, and IL-1β) by collecting and processing clinical data and serum samples from the experimental group, and then combining them with target cDNA to quantitatively detect the concentration of biomarkers. Through enzyme-linked immunosorbent assay (ELISA), antigen-antibody specific binding and enzymatic signal amplification are used to ensure high-sensitivity detection of biomarker concentrations. At the same time, by comparing the absorbance with a standard curve, the concentration of biomarkers can be accurately quantified. This process ensures the stability and repeatability of the data and improves the accuracy of the analysis, especially in the individualized assessment of asthma risk.

[0135] S4, constructing a bronchial asthma prediction model and outputting a predicted risk probability value of asthma for each patient;

[0136] Specifically, constructing a bronchial asthma prediction model and outputting a predicted risk probability value of each patient suffering from asthma includes: integrating the clinical data collected in the experimental group and the extracted target cDNA and biomarker concentrations into a data set;

[0137] Divide the dataset into training and testing sets;

[0138] Use CatBoost as the classifier model to build a bronchial asthma prediction model and set hyperparameters;

[0139] The setting of hyperparameters includes setting the number of iterations, the learning rate, and the depth of the decision tree;

[0140] The training set was input into the bronchial asthma prediction model. The training process included multiple iterations. Multi-classification cross entropy loss was selected as the loss function. The loss function was optimized by gradient descent in each iteration.

[0141] After the bronchial asthma prediction model training is completed, the performance of the model is evaluated using the test set;

[0142] Use cross-validation to verify the impact of different hyperparameter combinations on model performance and select the optimal hyperparameters;

[0143] The bronchial asthma prediction model is used to calculate all the patient's data through each decision tree to obtain the initial predicted probability of asthma risk M. i , the formula is:

[0144]

[0145] in, represents the initial predicted probability of asthma risk output by the CatBoost model for the i-th patient, represents the total number of decision trees, represents the output of the nth decision tree for all the data of the i-th patient (each decision tree makes a predicted output value based on the data of the i-th patient), represents the input data of the i-th patient, Indicates the weight of the nth decision tree. The CatBoost model automatically generates these weights during the training process based on the performance of each decision tree. The higher the weight, the greater the contribution of the decision tree to the final prediction result.

[0146] Based on the target cDNA extracted from the data set, the gene expression part is constructed to capture the risk signals of patients at the genetic level. The formula is:

[0147]

[0148] represents the contribution of gene expression to the asthma risk of the i-th patient, represents the relative expression level of NLRP3 in the i-th patient, represents the relative expression of IL-1β in the i-th patient, represents the relative expression level of Caspase-1 in the i-th patient;

[0149] The numerator represents the relative expression of NLRP3 and IL-1β genes, reflecting the key role of these genes in asthma pathology. The denominator represents the expression of Caspase-1, which is used to adjust the impact of overall gene expression and prevent the denominator from being zero. The logarithmic function is introduced to smooth the extreme gene expression values ​​and avoid prediction bias caused by abnormal data.

[0150] In traditional models, gene expression is usually weighted and summed in a linear manner, or the expression level of a gene is used alone as an indicator of risk assessment. This method lacks consideration of the complex interactions between genes, especially the interaction and regulatory relationship between genes in the pathological process. This formula captures the interaction between these genes in asthma inflammation by introducing the ratio relationship of NLRP3, IL-1β and Caspase-1 genes. In particular, through fractional and logarithmic operations, the influence of extreme values ​​is further smoothed, making the interaction between gene expressions more consistent with actual biological responses. Existing linear models usually find it difficult to capture these complex nonlinear relationships.

[0151] Using the biomarker concentrations in the dataset, the biomarker concentration part was constructed using the formula:

[0152]

[0153] in, represents the impact of the biomarker concentration on the risk prediction of the i-th patient, represents the biomarker concentration of TNF-α of the i-th patient, represents the biomarker concentration of IL-6 of the i-th patient, represents the biomarker concentration of IL-1β of the i-th patient;

[0154] The numerator part represents the product of TNF-α and IL-6 concentrations, reflecting their synergistic effect in the inflammatory response, and the denominator part represents the concentration of IL-1β, which regulates the effect of the overall marker on the inflammatory response and prevents the formula from becoming invalid when the concentration is too low;

[0155] In conventional biomarker assessments, many studies only consider the independent effects of a single marker, and usually use the absolute concentration of the marker to determine the patient's asthma risk. However, inflammatory markers are often the result of the synergistic effects of multiple molecules, and simple univariate analysis tends to ignore the interactive effects of these markers. This formula explicitly reflects the synergistic effects of these two important inflammatory markers by multiplying the concentrations of TNF-α and IL-6. In addition, IL-1β appears in the denominator as a regulatory molecule, which reflects the inhibitory effect of IL-1β concentration on the overall inflammatory response in certain cases by adjusting the product result.

[0156] The clinical data part is constructed using the patient's clinical data. The formula is:

[0157]

[0158] in, represents the contribution of the clinical data to the asthma risk of the i-th patient, represents the allergy history of the i-th patient, represents the family asthma history of the i-th patient, represents the medical history record of the i-th patient;

[0159] By multiplying the allergy history and family history, the impact of patients having dual risk factors is enhanced, and a logarithmic function is introduced to smooth the product results between data to prevent excessive values ​​from causing prediction instability;

[0160] In existing asthma risk prediction models, clinical data (such as allergy history, family history, and asthma history) are usually simply processed in a weighted addition manner, that is, the impact of each data is directly added together, lacking consideration of more complex data interactions. In addition, in traditional models, the nonlinear contribution of these data is not fully reflected. This formula captures the interaction of these two factors by multiplying the allergy history and family history. In particular, when a patient has both, the risk of asthma will be significantly increased. The logarithmic operation smoothes the impact of clinical data and avoids the impact of too large or too small values ​​when directly adding.

[0161] Construct the final risk probability formula to predict the patient's asthma risk probability. The formula is:

[0162]

[0163] in, represents the final predicted probability value of asthma risk for the i-th patient, Represents the Sigmoid function, which is used to compress the input value between 0 and 1 to generate a probability value;

[0164] According to the output predicted risk probability value of each patient suffering from asthma, a low risk probability threshold T is set. A With high risk probability threshold T B , classifying the predicted risk probability values ​​based on the risk probability threshold;

[0165] like , indicating that the patient is a low-risk asthma patient;

[0166] like , indicating that the patient is a medium-risk asthma patient;

[0167] like , indicating that the patient is a high-risk asthma patient;

[0168] The classification of low-risk asthma patients, medium-risk asthma patients, and high-risk asthma patients was ranked from low to high;

[0169] Each patient is sampled regularly and integrated into a new data set to input into the bronchial asthma prediction model, which outputs a new predicted risk probability value.

[0170] Through multi-dimensional data integration and complex nonlinear model processing, the accuracy and precision of bronchial asthma risk prediction are significantly improved. The present invention captures the interaction of asthma-related genes in inflammation by introducing gene expression levels. By constructing fractions and logarithmic processing, the problem that the linear weighted method in the prior art cannot capture the complex interaction between genes is solved, and the prediction bias caused by extreme gene expression values ​​is avoided, ensuring that the key regulatory role of genes in asthma pathology is reasonably expressed. The biomarker concentration part further reflects the synergistic effect of these markers in the inflammatory process, and uses IL-1β as a regulatory factor to balance the interaction effects between markers. This multi-marker synergistic processing method overcomes the limitation of only considering a single marker in the traditional model, avoids the shortcomings of simple absolute concentration evaluation, and enables a comprehensive evaluation of the overall picture of the inflammatory response. The clinical data part strengthens the influence of multiple risk factors through nonlinear processing. In traditional models, clinical data are often processed with simple weighting, ignoring the compound impact of multiple factors on risk when they exist at the same time. The present invention enhances the interaction effect between data by introducing the product operation of allergy history and family history, combined with logarithmic smoothing processing, especially when the patient has multiple risk factors at the same time, which significantly improves the model's ability to identify high-risk asthma patients. The introduced CatBoost model can capture complex nonlinear relationships and high-dimensional features through multiple decision trees and iterative optimization, and combines the multi-classification cross entropy loss function for model optimization, which significantly improves the accuracy of asthma risk prediction. The application of incremental learning method enables the model to dynamically update hyperparameters and adapt to new data input at any time, ensuring that the model has continuous self-optimization capabilities to meet real-time prediction needs.

[0171] S5, stores and manages all data in a relational database;

[0172] Specifically, storing and managing all data in a relational database includes selecting a relational database to store and manage patient data sets and their analysis results, designing a database table structure to store different types of data, setting up regular backup tasks, backing up all data in the database, managing permissions for database users, and encrypting and storing sensitive data.

[0173] By storing and managing all data in a relational database, it is possible to achieve systematic storage of patient data sets and analysis results, ensure data integrity and consistency, design database table structures to store different types of data so that data is easy to query and analyze, and regular backup tasks can effectively prevent data loss. Access control of database users is performed through permission management to ensure data security. At the same time, sensitive information is encrypted and stored to further improve data privacy and security. This process helps to optimize data management and analysis processes and provide reliable data support for personalized medical decisions. Example

[0174] Reference Figure 3 , which is the second embodiment of the present invention, and which is different from the previous embodiment, provides a bronchial asthma prediction system, including:

[0175] RNA extraction module, used to define the experimental group and the control group and extract RNA from the experimental group and the control group for cryopreservation;

[0176] cDNA selection module, used to perform quantitative PCR analysis on cryopreserved RNA and extract target cDNA;

[0177] The concentration calculation module is used to collect clinical data of the experimental group and calculate the corresponding biomarker concentration based on the target cDNA;

[0178] A model building module is used to build a bronchial asthma prediction model and output a predicted risk probability value of asthma for each patient;

[0179] The data storage module is used to store and manage all data in a relational database.

[0180] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0181] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0182] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0183] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit with a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit with a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0184] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for predicting bronchial asthma, characterized in that: include: The experimental and control groups were defined and RNA was extracted from the experimental and control groups for cryopreservation; The cryopreserved RNA was subjected to quantitative PCR analysis and target cDNA was extracted; Collect clinical data of the experimental group and calculate the corresponding biomarker concentration based on the target cDNA; Construct a bronchial asthma prediction model and output the predicted risk probability value of asthma for each patient; Store and manage all data in a relational database; The method of constructing a bronchial asthma prediction model and outputting a predicted risk probability value of each patient suffering from asthma includes: integrating the clinical data collected in the experimental group and the extracted target cDNA and biomarker concentration into a data set; Divide the dataset into training and testing sets; Use CatBoost as the classifier model to build a bronchial asthma prediction model and set hyperparameters; The setting of hyperparameters includes setting the number of iterations, the learning rate, and the depth of the decision tree; The training set was input into the bronchial asthma prediction model. The training process included multiple iterations. Multi-classification cross entropy loss was selected as the loss function. The loss function was optimized by gradient descent in each iteration. After the bronchial asthma prediction model training is completed, the performance of the model is evaluated using the test set; Use cross-validation to verify the impact of different hyperparameter combinations on model performance and select the optimal hyperparameters; The bronchial asthma prediction model is used to calculate all the patient's data through each decision tree to obtain the initial predicted probability of asthma risk M. i ; Based on the target cDNA extracted from the data set, the gene expression part is constructed to capture the risk signals of patients at the genetic level. The formula is: in, represents the contribution of gene expression to the asthma risk of the i-th patient, represents the relative expression level of NLRP3 in the i-th patient, represents the relative expression of IL-1β in the i-th patient, represents the relative expression level of Caspase-1 in the i-th patient; Using the biomarker concentrations in the dataset, the biomarker concentration part was constructed using the formula: in, represents the impact of the biomarker concentration on the risk prediction of the i-th patient, represents the biomarker concentration of TNF-α of the i-th patient, represents the biomarker concentration of IL-6 of the i-th patient, represents the biomarker concentration of IL-1β of the i-th patient; The clinical data part is constructed using the patient's clinical data. The formula is: in, represents the contribution of the clinical data to the asthma risk of the i-th patient, represents the allergy history of the i-th patient, represents the family asthma history of the i-th patient, represents the medical history record of the i-th patient; Construct the final risk probability formula to predict the patient's asthma risk probability. The formula is: in, represents the final predicted probability value of asthma risk for the i-th patient, Represents the Sigmoid function; According to the output predicted risk probability value of each patient suffering from asthma, a low risk probability threshold T is set. A With high risk probability threshold T B , classifying the predicted risk probability values ​​based on the risk probability threshold; like , indicating that the patient is a low-risk asthma patient; like , indicating that the patient is a medium-risk asthma patient; like , indicating that the patient is a high-risk asthma patient; The classification of low-risk asthma patients, medium-risk asthma patients, and high-risk asthma patients was ranked from low to high; Each patient is sampled regularly and integrated into a new data set to be input into the bronchial asthma prediction model to output a new predicted risk probability value.

2. The bronchial asthma prediction method according to claim 1, characterized in that: The step of defining the experimental group and the control group and extracting RNA from the experimental group and the control group for cryopreservation comprises: defining the experimental group and the control group according to the experimental requirements; The experimental group refers to patients who were judged to be suspected of asthma; The control group refers to healthy individuals; The nasal cavity, pharynx, and airway of the members of the experimental and control groups were locally anesthetized with 1% lidocaine, and a fiberoptic bronchoscope was slowly inserted into the airway; Through a fiberoptic bronchoscope, a disposable cell brush is inserted into the bronchus to collect epithelial cell samples on the surface of the bronchial epithelium; The collected epithelial cell samples are placed in a test tube containing a preservation solution, and RNA in the epithelial cell samples is extracted using an RNA extraction kit; The purity and concentration of RNA were tested using a spectrophotometer; The detection of RNA purity and concentration includes using a spectrophotometer to evaluate RNA purity through an absorbance ratio, and according to the purity evaluation standard, setting the normal range of the absorbance ratio to be between 1.8 and 2.

1. If the absorbance ratio of the RNA is not within the normal range, it means that the RNA purity does not meet the standard, and then using an ethanol precipitation method to re-purify the RNA. If the absorbance ratio of the RNA is within the normal range, it means that the RNA purity meets the standard, and then performing a concentration test; According to the experimental requirements, the normal concentration range of RNA is set. If the concentration of RNA after the purity test is not within the normal concentration range of RNA, it means that the RNA concentration does not meet the standard, then the RNA sample is concentrated and the RNA is re-extracted. If the concentration of RNA after the purity test is within the normal concentration range of RNA, it means that the RNA concentration meets the standard, then low-temperature storage treatment is performed.

3. The bronchial asthma prediction method according to claim 2, characterized in that: The step of performing quantitative PCR analysis on the cryopreserved RNA and extracting target cDNA includes using a reverse transcription kit to perform a reverse transcription reaction from RNA to cDNA, and obtaining cDNA for quantitative PCR analysis after the reverse transcription reaction is completed; A single reaction system suitable for quantitative PCR analysis was configured, and SYBR Green method was used as the detection method; The configuration of the single reaction system includes configuring SYBR Green Master Mix premix reagent, template cDNA, forward primer and reverse primer; The cDNA targeted by the forward primer and the reverse primer in a single reaction system is designed and called the target cDNA; The target cDNAs include NLRP3, IL-1β, and Caspase-1; After the single reaction system is configured, the target cDNA is added as a template into the single reaction system to obtain a configured qPCR reaction mixture, and the qPCR reaction mixture is placed in a qPCR reaction plate and sealed with a transparent optical sealing film; Place the sealed reaction plate into a real-time fluorescence quantitative qPCR instrument, set the reaction program and perform qPCR amplification reaction; The reaction program includes initial deformation, cyclic reaction and dissolution curve; A predetermined threshold is set by the intensity of the fluorescence signal, the number of cycles that first exceed the predetermined threshold in each cycle of the experimental group and the control group is monitored, and the number of cycles that first exceed the predetermined threshold is taken as the CT value; According to the obtained CT value, the CT value difference between the target cDNA and the internal reference cDNA in the experimental group and the control group was calculated respectively; According to the obtained CT value difference between the target cDNA and the internal reference cDNA, the difference in CT value between the experimental group and the control group was calculated; The relative expression levels of the target cDNA in the experimental group and the control group were calculated.

4. The bronchial asthma prediction method according to claim 3, characterized in that: The collecting clinical data of the experimental group and calculating the corresponding biomarker concentration according to the target cDNA includes: collecting the clinical data of the experimental group from the electronic medical record system of the hospital; The clinical data included allergy history, family asthma history, and medical history records; If you have a history of allergies, please record the history of allergies. g,i is 1, otherwise it is 0; If there is a family history of asthma, please record the family history of asthma. j,i is 1, otherwise it is 0; If you have a history of asthma, please record the medical history. z,i is 1, otherwise it is 0; Collect serum samples from the experimental group and process the samples. Dilute the processed serum samples in proportion according to experimental needs and add them to a 96-well plate. Based on the target cDNA, the absorbance value of the biomarker in the serum sample was detected by enzyme-linked immunosorbent assay, and the concentration of each biomarker was calculated by comparing the absorbance with the standard curve of standards with known concentrations; The biomarker concentrations include the concentrations of TNF-α, IL-6, and IL-1β in serum.

5. The bronchial asthma prediction method according to claim 4, characterized in that: The storing and managing of all data in a relational database includes selecting a relational database to store and manage patient data sets and analysis results, designing a database table structure to store different types of data, setting regular backup tasks, backing up all data in the database, managing permissions for database users, and encrypting and storing sensitive data.

6. A bronchial asthma prediction system based on the bronchial asthma prediction method according to any one of claims 1 to 5, characterized in that: include, RNA extraction module, used to define the experimental group and the control group and extract RNA from the experimental group and the control group respectively for cryopreservation; cDNA selection module, used to perform quantitative PCR analysis on cryopreserved RNA and extract target cDNA; The concentration calculation module is used to collect clinical data of the experimental group and calculate the corresponding biomarker concentration based on the target cDNA; A model building module, used to build a bronchial asthma prediction model and output a predicted risk probability value of asthma for each patient; The data storage module is used to store and manage all data in a relational database.

Images