Kit for detecting expression quantity of HGFAC protein in body fluid, construction method of IBD detection model and application of method in IBD auxiliary diagnosis, treatment effect monitoring and prognosis outcome prediction

Abstract

The invention provides a detection kit for the expression quantity of HGFAC protein in body fluid, a construction method of an IBD detection model and application of the method in IBD auxiliary diagnosis, treatment effect monitoring and prognosis outcome prediction, belongs to the field of biological medicine, and solves the problem of how to carry out IBD auxiliary diagnosis, treatment effect monitoring and prognosis outcome prediction through body fluid biomarkers. The method comprises the following steps: inputting body fluid data obtained by the HGFAC protein expression quantity in a body fluid sample detected by the detection kit into a Logistic regression model to obtain an IBD auxiliary diagnosis model; inputting target body fluid data after conventional treatment into a Logistic regression model to obtain an IBD treatment effect monitoring and prognosis outcome prediction model; and selecting an optimal random seed number according to an AUC value of an ROC curve established by the model to obtain an optimal model, and performing k-folding cross validation on the optimal model to obtain an optimal model. According to the invention, IBD auxiliary diagnosis, treatment effect monitoring and prognosis outcome prediction are carried out through the optimal model based on the body fluid biomarkers, and clinical requirements are met.

Description

Technical Field

[0001] The present invention belongs to the field of biomedicine technology, and in particular relates to a kit for detecting the expression of HGFAC protein in body fluids, a method for constructing an IBD detection model, and applications of the method in auxiliary diagnosis, treatment effect monitoring, and prognosis prediction of IBD. Background Art

[0002] Inflammatory bowel disease (IBD) is a chronic nonspecific inflammatory disease with unknown etiology, and its etiology has not yet been fully elucidated. It is divided into Crohn's disease (CD) and ulcerative colitis (UC). At present, the pathogenesis of IBD is not fully understood, but it is likely to involve multiple factors such as genetic susceptibility, immunity, intestinal microbiology, and environmental psychological factors. IBD not only affects the gastrointestinal tract, but also often leads to a series of other immune disorders such as psoriasis, which leads to a worse prognosis and outcome of patients, making treatment more challenging.

[0003] The importance of body fluid proteomics in IBD research is increasingly recognized. Body fluid-based markers are crucial for studying the occurrence and prognosis of IBD. Although this test method is predicted to provide simple and cost-effective diagnosis, it has been shelved due to the lack of ideal body fluid biomarkers.

[0004] Human hepatocyte growth factor activator (HGFAC) protein can activate human hepatocyte growth factor (HGF), which is mainly secreted into the circulation by the liver to stimulate lipid metabolism in the blood.

[0005] The inventors found that IBD patients have higher triglyceride levels, and CD patients have higher triglyceride levels than UC patients. This increase may be due to intestinal inflammation affecting lipid metabolism. Increased triglyceride levels are associated with the severity of the disease. Abnormal blood lipid levels may have a two-way relationship with IBD, aggravating each other and forming a vicious cycle. In addition to blood lipid levels, abnormal fat distribution is also closely related to the prognosis of CD. The visceral fat content of CD patients is higher than that of the healthy control group, forming a fat crawling sign near the intestine, and is associated with an increased risk of complex CD phenotypes, which is also a unique feature of IBD compared with other immune disorders. At the same time, Mendelian analysis shows that circulating HGFAC protein is closely related to the occurrence of IBD, and HGFAC levels are inversely proportional to the risk of IBD. Other studies have found that knocking out HGFAC impairs the repair function of intestinal epithelial cells, leading to weakened barrier function and increased susceptibility to IBD. Therefore, HGFAC can be used as a humoral marker for the diagnosis, monitoring of treatment effects, and prediction of prognosis of IBD.

[0006] In summary, the present invention uses machine learning to develop a kit for detecting the expression of HGFAC protein in body fluids, a method for constructing an IBD detection model, and the application of the method in auxiliary diagnosis, treatment effect monitoring, and prognosis prediction of IBD. The method is used for auxiliary diagnosis, treatment effect monitoring, and prognosis prediction of IBD, is suitable for clinical needs, is simple to operate, reduces the occurrence of medical accidents, and can save medical expenses for IBD patients, reduce pain during detection, and improve the quality of life of IBD patients. Summary of the invention

[0007] The body fluid detection method provides a simple and cost-effective diagnostic option. In order to solve the technical problem that the lack of ideal body fluid biomarkers in the clinic cannot be used for IBD auxiliary diagnosis, treatment effect monitoring and prognosis prediction, the present invention provides a body fluid HGFAC protein expression detection kit, an IBD detection model construction method and the application of the method in IBD auxiliary diagnosis, treatment effect monitoring and prognosis prediction. The present invention constructs an optimal IBD auxiliary diagnosis model and an optimal IBD treatment effect monitoring and prognosis prediction model through logistic regression, and determines whether the patient has the risk of IBD, the response to conventional treatment, and the prognosis and outcome of the patient through the optimal model; the present invention can be performed in a home or primary care environment through non-invasive detection, simple and fast operation, and can be performed through the optimal model constructed to help predict the occurrence and development of IBD, the treatment effect and prognosis of IBD in advance, and to perform risk prediction and early intervention before the IBD patient further worsens.

[0008] In order to achieve the above object, the present invention is specifically implemented through the following technical solutions:

[0009] The present invention provides a kit for detecting the expression of HGFAC protein in body fluids, comprising a rabbit polyclonal antibody against HGFAC and horseradish peroxidase HRP;

[0010] Collecting a body fluid sample; the body fluid sample includes a serum sample and / or a urine sample;

[0011] The rabbit polyclonal antibody against HGFAC binds to the HGFAC protein in the body fluid sample;

[0012] Horseradish peroxidase HRP was added to label the rabbit polyclonal antibody against HGFAC, and the expression level of HGFAC protein in the body fluid samples was detected.

[0013] Furthermore, the method for collecting serum samples from body fluid samples includes:

[0014] Use a vacuum tube to collect a 5-10 mL morning fasting venous blood sample for 8-12 hours;

[0015] The venous blood sample was allowed to stand at room temperature until the blood coagulated, and then a test tube free of pyrogens and endotoxins was used. Any cell stimulation was avoided during the operation. After the blood was collected, it was centrifuged at 5000 rpm for 10 minutes to separate the serum and red blood cells in the blood and frozen for later use to obtain a serum sample.

[0016] Furthermore, the method for collecting urine samples from body fluid samples includes:

[0017] The patient has not urinated for at least 2 hours, and a sterile urine collection container is used to collect 40-50 ml of urine during the patient's first urination in the morning, and the urine sample is immediately placed in a -80°C refrigerator for storage to obtain a urine sample. The present invention also provides a method for constructing an IBD detection model for the expression of HGFAC protein in body fluids, including a kit for detecting the expression of HGFAC protein in body fluids, comprising:

[0018] Step 1: Binding the HGFAC protein expression level in a body fluid sample detected by a body fluid HGFAC protein expression detection kit with corresponding clinical data to obtain body fluid data, and establishing an IBD body fluid proteomics database through the body fluid data; the body fluid data includes serum data and urine data;

[0019] Step 2: Use Jupyter software to import the linear model library, set the random seed and tolerance error to initialize the Logistic regression model; input the serum data and urine data in the IBD body fluid proteomics database into the Logistic regression model for fitting, and output the coefficients and intercepts to obtain the IBD auxiliary diagnosis model;

[0020] Step 3. If the auxiliary diagnosis result of the IBD auxiliary diagnosis model indicates that there is a risk of IBD, the target body fluid data of the endoscopic score SES-CD before and after conventional treatment that decreases by more than the set threshold is used as the response; the target serum data after conventional treatment is input into the Logistic regression model for fitting, and the coefficients and intercepts are output to obtain the IBD treatment effect monitoring and prognosis prediction model of the serum data; or

[0021] Input the target urine data after conventional treatment into the Logistic regression model for fitting, and output the coefficients and intercept to obtain the IBD treatment effect monitoring and prognosis prediction model of urine data; or

[0022] The target serum data and target urine data after conventional treatment were simultaneously input into the Logistic regression model for fitting, and the coefficients and intercepts were output to obtain the IBD treatment effect monitoring and prognosis prediction model combining serum data and urine data;

[0023] Step 4: Establish ROC curves based on the constructed IBD auxiliary diagnosis model and IBD treatment effect monitoring and prognosis prediction model, respectively. Select the optimal number of random seeds according to the AUC value to obtain the optimal model. Use k-fold cross validation to obtain the optimal model to obtain the best IBD auxiliary diagnosis model and the best IBD treatment effect monitoring and prognosis prediction model.

[0024] In step 1, the body fluid data in the IBD body fluid proteomics database are the body fluid data for which the HGFAC protein expression level in a body fluid detected by the HGFAC protein expression detection kit is less than 3 times the interquartile range and the body fluid data for which the clinical data corresponding to the HGFAC protein expression level is missing are screened out.

[0025] In step 2, the IBD auxiliary diagnosis model is:

[0026] Y1=a0+a1*x1+a2*y1;

[0027] Among them, Y1 is the auxiliary diagnosis score of IBD. If Y1≥0, the auxiliary diagnosis result is no risk of IBD. If Y1<0, the auxiliary diagnosis result indicates the risk of IBD.

[0028] a0 is the intercept of the IBD auxiliary diagnosis model, which is a constant term and represents the predicted value of the IBD auxiliary diagnosis model when all body fluid data are 0;

[0029] a1 is the regression coefficient of serum data;

[0030] a2 is the regression coefficient of urine data;

[0031] x1 is serum data;

[0032] y1 is urine data.

[0033] In step 3, the IBD treatment effect monitoring and prognosis prediction model of serum data is:

[0034] Y2=b0+b1*x2;

[0035] Among them, Y2 is the IBD treatment effect monitoring and prognosis prediction score of serum data. If Y2≥0, it indicates that the patient has a good response to conventional treatment and a good prognosis. If Y2<0, it indicates that the treatment response is not good, the prognosis is poor, and there is a risk of secondary other immune disorders.

[0036] b0 is the intercept of the IBD treatment effect monitoring and prognosis prediction model of serum data, which is a constant term, indicating the predicted value of the IBD treatment effect monitoring and prognosis prediction model of serum data when the target serum data is 0;

[0037] b1 is the regression coefficient of the target serum data;

[0038] x2 is the target serum data.

[0039] In step 3, the IBD treatment effect monitoring and prognosis prediction model of urine data is:

[0040] Y3=c0+c1*y2;

[0041] Among them, Y3 is the IBD treatment effect monitoring and prognosis prediction score of urine data. If Y3≥0, it indicates that the patient has a good response to conventional treatment and a good prognosis. If Y3<0, it indicates that the treatment response is not good, the prognosis is poor, and there is a risk of secondary other immune disorders.

[0042] c0 is the intercept of the IBD treatment effect monitoring and prognosis prediction model of urine data, which is a constant term and represents the predicted value of the IBD treatment effect monitoring and prognosis prediction model of urine data when the target urine data is 0;

[0043] c1 is the regression coefficient of the target urine data;

[0044] y2 is the target urine data.

[0045] In step 3, the IBD treatment effect monitoring and prognosis prediction model based on serum and urine data is:

[0046] Y4=d0+d1*x2+d2*y2;

[0047] Among them, Y4 is the IBD treatment effect monitoring and prognosis prediction score combined with serum data and urine data. If Y4≥0, it indicates that the patient has a good response to conventional treatment and a good prognosis. If Y4<0, it indicates that the treatment response is not good, the prognosis is poor, and there is a risk of secondary other immune disorders.

[0048] d0 is the intercept of the IBD treatment effect monitoring and prognosis prediction model based on the combined serum data and urine data, which is a constant term, indicating the predicted value of the IBD treatment effect monitoring and prognosis prediction model based on the combined serum data and urine data when all target body fluid data are 0;

[0049] d1 is the joint regression coefficient of target serum data;

[0050] d2 is the joint regression coefficient of the target urine data;

[0051] x2 is the target serum data;

[0052] y2 is the target urine data.

[0053] The present invention also provides an IBD detection model construction method for HGFAC protein expression in body fluids and its application in IBD auxiliary diagnosis, treatment effect monitoring and prognosis prediction.

[0054] The beneficial effects of the present invention are:

[0055] The present invention provides a kit for detecting the expression of HGFAC protein in body fluids, a method for constructing an IBD detection model, and the application of the method in auxiliary diagnosis, treatment effect monitoring, and prognosis prediction of IBD. The method is non-invasive, simple, and rapid to operate, can be performed at home or in a primary care setting, and is helpful for early prediction of the occurrence and development of IBD, treatment effect, and prognosis, and for risk prediction and early intervention of IBD patients before further exacerbation.

[0056] Compared with the existing technology, the present invention provides a minimally invasive or even non-invasive and cost-effective method for auxiliary diagnosis, treatment effect monitoring and prognosis prediction of IBD, when current research still lacks ideal body fluid biomarkers to diagnose and predict the occurrence and development of IBD. Compared with traditional examination methods such as ordinary colonoscopy, capsule endoscopy, CT and MRI, body fluid testing is more convenient and quick, minimally invasive or even non-invasive, cheaper, and more acceptable to patients. Since the main function of HGFAC is to regulate lipid metabolism and is closely related to the inflammatory level of IBD, and an important feature of IBD patients is lipid metabolism disorder, a large amount of fat accumulates in the intestine to form fat crawling sign, which means that the present invention can detect the expression of HGFAC and can not only be used for the diagnosis, treatment effect monitoring and prognosis prediction of IBD, but also can effectively perform differential diagnosis with other inflammatory and immune diseases. It is suitable for clinical needs, simple to operate, and reduces the occurrence of medical accidents. At the same time, it can save medical expenses for IBD patients, reduce pain during testing, control the development of the disease, and prevent the occurrence of the disease. It provides an important tool for auxiliary diagnosis, treatment effect monitoring, prognosis judgment and management of complications of IBD, which helps to improve the quality of life and treatment effect of patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The present invention is further described in detail below based on the accompanying drawings and embodiments.

[0058] Figure 1 A kit for detecting HGFAC protein expression in body fluids provided by the present invention detects the expression of HGFAC protein in serum samples in body fluids of HC (healthy control) and IBD patients.

[0059] Figure 2 A kit for detecting the expression of HGFAC protein in body fluids provided by the present invention is used to detect the expression of HGFAC protein in urine samples in body fluids of HC (healthy control) and IBD patients.

[0060] Figure 3 To analyze the correlation between HGFAC protein expression and IBD clinical data and the consistency of HGFAC protein expression in serum and urine samples using bioinformatics.

[0061] Figure 4 The AUC value of the optimal IBD auxiliary diagnosis model showed that HGFAC in body fluids could predict the risk of IBD.

[0062] Figure 5 The AUC value of the best prognosis prediction model shows the predictive efficacy of HGFAC in body fluids in predicting the treatment effect and prognosis of IBD patients. DETAILED DESCRIPTION

[0063] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0064] Since IBD patients generally have abnormal blood lipid levels and fat distribution, which is also the key to distinguishing IBD from other immune disorders and inflammatory diseases, differential analysis and functional enrichment analysis were used to select proteins with the most obvious changes in body fluid samples and those that are obviously related to lipid metabolism. The present invention ultimately identified HGFAC as a body fluid marker for IBD.

[0065] The embodiment of the present invention provides a kit for detecting the expression of HGFAC protein in body fluids, comprising a rabbit polyclonal antibody against HGFAC and horseradish peroxidase HRP;

[0066] Collecting a body fluid sample; the body fluid sample includes a serum sample and / or a urine sample;

[0067] The rabbit polyclonal antibody against HGFAC binds to the HGFAC protein in the body fluid sample;

[0068] Horseradish peroxidase HRP was added to label the rabbit polyclonal antibody against HGFAC, and the expression level of HGFAC protein in the body fluid samples was detected.

[0069] Among them, rabbit polyclonal antibodies against HGFAC include but are not limited to: Affinity#DF9570 Rabbit polyclonal antibody to HGFAC; Acmec#AC50820Anti-HGFAC Polyclonal Antibody; Servicebio#GB113116Anti-HGFAC Rabbit pAb, etc.

[0070] Proteomic data analysis: The HGFAC protein expression in body fluid samples obtained by a body fluid HGFAC protein expression detection kit was statistically analyzed using R software (v4.4.0) to obtain the expression differences between healthy controls and IBD patients. Figure 1 and Figure 2 As shown, Figure 1 and Figure 2The quantitative expression of HGFAC protein in serum and urine samples is shown. The results show that the expression of HGFAC protein in serum and urine samples of IBD patients is significantly reduced, where the y-axis is the specific expression of the protein and the asterisk represents a statistically significant difference.

[0071] Among them, the collection method of serum samples in body fluid samples includes:

[0072] Use a vacuum tube to collect a 5-10 mL morning fasting venous blood sample for 8-12 hours;

[0073] The venous blood sample was allowed to stand at room temperature until the blood coagulated, and then a test tube free of pyrogens and endotoxins was used. Any cell stimulation was avoided during the operation. After the blood was collected, it was centrifuged at 5000 rpm for 10 minutes to separate the serum and red blood cells in the blood and frozen for later use to obtain a serum sample.

[0074] For example: It is recommended that patients fast for 8-12 hours in the morning to collect blood. Use a vacuum blood collection tube to collect about 5-10mL of venous blood samples. After collecting the blood sample, let it stand at room temperature for 30 minutes to allow the blood to coagulate. Use a test tube that does not contain pyrogens and endotoxins, and avoid any cell stimulation during the operation. After collecting the blood, use a centrifuge at 5000 rpm for 10 minutes (4°C) to separate the serum. Take the supernatant (serum), divide it into low-adsorption centrifuge tubes, and immediately freeze it at -80°C.

[0075] Among them, the collection methods of urine samples in body fluid samples include:

[0076] The patient has not urinated for at least 2 hours. Use a sterile urine collection container to collect 40-50 ml of urine when the patient urinates for the first time in the morning, and immediately put it in a -80℃ refrigerator for later use to obtain a urine sample.

[0077] For example, patients should collect urine samples during the first urination in the morning. It is recommended that patients should not urinate at least 2 hours before urine collection to ensure the representativeness of the sample and the concentration of biomarkers. Use a sterile urine collection container to collect approximately 40-50 ml of urine. Then immediately store it in a -80℃ refrigerator for later use.

[0078] The present invention also provides a method for constructing an IBD detection model for HGFAC protein expression in body fluids, comprising a kit for detecting HGFAC protein expression in body fluids, comprising:

[0079] Step 1: Binding the HGFAC protein expression level in a body fluid sample detected by a body fluid HGFAC protein expression detection kit with corresponding clinical data to obtain body fluid data, and establishing an IBD body fluid proteomics database through the body fluid data; the body fluid data includes serum data and urine data;

[0080] Based on clinical data, the Pearson correlation score between HGFAC protein expression and IBD inflammatory indicators, endoscopic scores, drug response rates, and the probability of secondary psoriasis was established. The Pearson correlation score analysis was performed by collecting clinical data corresponding to IBD patients and body fluid samples to verify the correlation analysis between HGFAC protein expression and patient-related inflammatory indicators commonly used in clinical data. Figure 3 As shown, Figure 3 The Pearson correlation between HGFAC protein expression and various clinical data of IBD patients was shown. The results showed that HGFAC protein expression was negatively correlated with inflammatory indicators in the clinical data of IBD patients, among which PS was whether there was secondary psoriasis, NEUT was neutrophil count, CRP was C-reactive protein, WBC was white blood cell count, ESR was erythrocyte sedimentation rate, SES-CD was endoscopic score, and TG was blood triglyceride; at the same time, the HGFAC protein expression in serum samples and urine samples showed a high degree of consistency, red represented positive correlation, blue represented negative correlation, the darker the color, the stronger the correlation, and the asterisk represented statistical difference.

[0081] Combination Figure 1 and Figure 2 The expression of HGFAC protein in Figure 3 We established an IBD body fluid proteomics database based on the clinical data in the literature, performed quality control on the proteomics data in the database, and statistically analyzed the individual protein identification quantities of all samples. We screened out body fluid data with an HGFAC protein expression quantity detection kit identification quantity lower than 3 times the interquartile range of HGFAC protein expression quantity and body fluid data with missing clinical data corresponding to HGFAC protein expression quantity.

[0082] Step 2: Use Jupyter software (Python 3.8 version) to import the linear model library, set the random seed and tolerance error (can be set to 1e-6) to initialize the Logistic regression model; input the serum data and urine data in the IBD body fluid proteomics database into the Logistic regression model for fitting, and output the coefficients and intercepts to obtain the IBD auxiliary diagnosis model;

[0083] Logistic regression is a classification algorithm based on the logistic function. Its basic idea is to map the results of linear regression to the [0,1] interval through the logistic function, thereby converting the output of any real number domain into a category probability with a probabilistic interpretation.

[0084] Step 3. If the auxiliary diagnosis result of the IBD auxiliary diagnosis model indicates that there is a risk of IBD, the target body fluid data of the endoscopic score SES-CD before and after conventional treatment that decreases by more than the set threshold is used as the response; the target serum data after conventional treatment is input into the Logistic regression model for fitting, and the coefficients and intercepts are output to obtain the IBD treatment effect monitoring and prognosis prediction model of the serum data; or

[0085] Input the target urine data after conventional treatment into the Logistic regression model for fitting, and output the coefficients and intercept to obtain the IBD treatment effect monitoring and prognosis prediction model of urine data; or

[0086] The target serum data and target urine data after conventional treatment were simultaneously input into the Logistic regression model for fitting, and the coefficients and intercepts were output to obtain the IBD treatment effect monitoring and prognosis prediction model combining serum data and urine data;

[0087] Step 4. ROC curves were established based on the constructed IBD auxiliary diagnosis model and IBD treatment effect monitoring and prognosis prediction model, and the optimal random seed number was selected according to the AUC value to obtain the optimal model. The optimal model was validated by k-fold cross validation in the training set and the test set to obtain the best IBD auxiliary diagnosis model and the best IBD treatment effect monitoring and prognosis prediction model to evaluate the accuracy of predicting the risk of IBD incidence, treatment effect and prognosis.

[0088] The AUC value is an important indicator to measure the prediction performance of the binary classification model. The AUC value range is 0.5-1; 0.5 means that the model performance is the same as random guessing, without any discrimination ability, and 1.0 means that the model perfectly predicts all positive and negative samples.

[0089] In step 2, the IBD auxiliary diagnosis model is:

[0090] Y1=a0+a1*x1+a2*y1;

[0091] Among them, Y1 is the auxiliary diagnosis score of IBD. If Y1≥0, the auxiliary diagnosis result is no risk of IBD. If Y1<0, the auxiliary diagnosis result indicates the risk of IBD.

[0092] a0 is the intercept of the IBD auxiliary diagnosis model, which is a constant term and represents the predicted value of the IBD auxiliary diagnosis model when all body fluid data are 0;

[0093] a1 is the regression coefficient of serum data;

[0094] a2 is the regression coefficient of urine data;

[0095] x1 is serum data;

[0096] y1 is urine data.

[0097] In step 3, the IBD treatment effect monitoring and prognosis prediction model of serum data is:

[0098] Y2=b0+b1*x2;

[0099] Among them, Y2 is the IBD treatment effect monitoring and prognosis prediction score of serum data. If Y2≥0, it indicates that the patient has a good response to conventional treatment and a good prognosis. If Y2<0, it indicates that the treatment response is not good, the prognosis is poor, and there is a risk of secondary other immune disorders.

[0100] b0 is the intercept of the IBD treatment effect monitoring and prognosis prediction model of serum data, which is a constant term, indicating the predicted value of the IBD treatment effect monitoring and prognosis prediction model of serum data when the target serum data is 0;

[0101] b1 is the regression coefficient of the target serum data;

[0102] x2 is the target serum data.

[0103] In step 3, the IBD treatment effect monitoring and prognosis prediction model of urine data is:

[0104] Y3=c0+c1*y2;

[0105] Among them, Y3 is the IBD treatment effect monitoring and prognosis prediction score of urine data. If Y3≥0, it indicates that the patient has a good response to conventional treatment and a good prognosis. If Y3<0, it indicates that the treatment response is not good, the prognosis is poor, and there is a risk of secondary other immune disorders.

[0106] c0 is the intercept of the IBD treatment effect monitoring and prognosis prediction model of urine data, which is a constant term and represents the predicted value of the IBD treatment effect monitoring and prognosis prediction model of urine data when the target urine data is 0;

[0107] c1 is the regression coefficient of the target urine data;

[0108] y2 is the target urine data.

[0109] In step 3, the IBD treatment effect monitoring and prognosis prediction model based on serum and urine data is:

[0110] Y4=d0+d1*x2+d2*y2;

[0111] Among them, Y4 is the IBD treatment effect monitoring and prognosis prediction score combined with serum data and urine data. If Y4≥0, it indicates that the patient has a good response to conventional treatment and a good prognosis. If Y4<0, it indicates that the treatment response is not good, the prognosis is poor, and there is a risk of secondary other immune disorders.

[0112] d0 is the intercept of the IBD treatment effect monitoring and prognosis prediction model based on the combined serum data and urine data, which is a constant term, indicating the predicted value of the IBD treatment effect monitoring and prognosis prediction model based on the combined serum data and urine data when all target body fluid data are 0;

[0113] d1 is the joint regression coefficient of target serum data;

[0114] d2 is the joint regression coefficient of the target urine data;

[0115] x2 is the target serum data;

[0116] y2 is the target urine data.

[0117] In order to illustrate the practicability of the optimal model of the present invention, a specific application example is now provided:

[0118] Combined with the existing clinical data and by entering the specific code: "from sklearn.linear_model importLogisticRegression; lr = LogisticRegression (random_state = 99, tol = 1e-6) # Logistic regression model; lr.fit (x, y); print ("Coef:", lr.coef_); print ("intercept", lr.intercept_)" to obtain the best IBD auxiliary diagnosis model and three parallel best IBD treatment effect monitoring and prognosis prediction models; among them,

[0119] The best IBD auxiliary diagnosis model is:

[0120] Y1=1.19611391+0.81323719*x1+0.72086323*y1.

[0121] Figure 4 The AUC value of the best IBD auxiliary diagnosis model for predicting the risk of IBD is shown. Since HGFAC is mainly secreted by the liver into the blood circulation to play a role, the embodiments of the present invention use the HGFAC protein expression in serum samples and urine samples for modeling, which is more representative.

[0122] The score is calculated according to the formula, such as Figure 4The result AUC = 0.863, showing that the HGFAC protein expression in serum data and urine data has a good predictive effect on the risk of IBD, indicating that the model can correctly determine the relative order of positive and negative samples in 86.3% of cases.

[0123] Next, this example predicts the treatment effect and prognosis of IBD patients, taking the decrease of endoscopic score SES-CD by more than 50% before and after conventional treatment as the response, and the IBD treatment effect monitoring and prognosis prediction model based on serum data is:

[0124] Y2=-1.40721159+1.14252761*x2

[0125] The IBD treatment effect monitoring and prognosis prediction model based on urine data is:

[0126] Y3=-1.32948032+1.13937744*y2

[0127] The IBD treatment effect monitoring and prognosis prediction model based on serum and urine data is as follows:

[0128] Y4=-2.10801654+0.95434642*x2+0.9606812*y2

[0129] Figure 5 The AUC value of the best IBD treatment effect monitoring and prognosis prediction model is shown. The score is calculated according to the formula, such as Figure 5 The result AUC = 0.80 shows that the HGFAC protein expression in serum data or urine data has a good predictive effect on the prognosis of IBD patients, and urine data is non-invasive, convenient and fast.

[0130] The combined prediction performance of serum data and urine data is better: AUC = 0.83. This indicates that by detecting the expression of HGFAC protein in serum data and urine data, it can be used for the diagnosis, treatment effect monitoring and prognosis prediction of IBD in a simple, efficient, minimally invasive or even non-invasive way; these three parallel models can be freely selected according to clinical needs.

[0131] The present invention also provides an IBD detection model construction method for HGFAC protein expression in body fluids and its application in IBD auxiliary diagnosis, treatment effect monitoring and prognosis prediction.

[0132] The beneficial effects of the present invention are:

[0133] The present invention provides a kit for detecting the expression of HGFAC protein in body fluids, a method for constructing an IBD detection model, and the application of the method in auxiliary diagnosis, treatment effect monitoring, and prognosis prediction of IBD. The method is non-invasive, simple, and rapid to operate, can be performed at home or in a primary care setting, and is helpful for early prediction of the occurrence and development of IBD, treatment effect, and prognosis, and for risk prediction and early intervention of IBD patients before further exacerbation.

[0134] Compared with the existing technology, the present invention provides a minimally invasive or even non-invasive and cost-effective method for auxiliary diagnosis, treatment effect monitoring and prognosis prediction of IBD, when current research still lacks ideal body fluid biomarkers to diagnose and predict the occurrence and development of IBD. Compared with traditional examination methods such as ordinary colonoscopy, capsule endoscopy, CT and MRI, body fluid testing is more convenient and quick, minimally invasive or even non-invasive, cheaper, and more acceptable to patients. Since the main function of HGFAC is to regulate lipid metabolism and is closely related to the inflammatory level of IBD, and an important feature of IBD patients is lipid metabolism disorder, a large amount of fat accumulates in the intestine to form fat crawling sign, which means that the present invention can detect the expression of HGFAC and can not only be used for the diagnosis, treatment effect monitoring and prognosis prediction of IBD, but also can effectively perform differential diagnosis with other inflammatory and immune diseases. It is suitable for clinical needs, simple to operate, and reduces the occurrence of medical accidents. At the same time, it can save medical expenses for IBD patients, reduce pain during testing, control the development of the disease, and prevent the occurrence of the disease. It provides an important tool for auxiliary diagnosis, treatment effect monitoring, prognosis judgment and management of complications of IBD, which helps to improve the quality of life and treatment effect of patients.

[0135] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A kit for detecting the expression of HGFAC protein in body fluids, characterized in that: Includes rabbit polyclonal antibody against HGFAC and horseradish peroxidase HRP; Collecting a body fluid sample; the body fluid sample includes a serum sample and / or a urine sample; The rabbit polyclonal antibody against HGFAC binds to the HGFAC protein in the body fluid sample; Horseradish peroxidase HRP was added to label the rabbit polyclonal antibody against HGFAC, and the expression level of HGFAC protein in the body fluid samples was detected.

2. A kit for detecting HGFAC protein expression in body fluids according to claim 1, characterized in that: The collection methods of serum samples among body fluid samples include: Use a vacuum tube to collect a 5-10 mL morning fasting venous blood sample for 8-12 hours; The venous blood sample was allowed to stand at room temperature until the blood coagulated, and then a test tube free of pyrogens and endotoxins was used. Any cell stimulation was avoided during the operation. After the blood was collected, it was centrifuged at 5000 rpm for 10 minutes to separate the serum and red blood cells in the blood and frozen for later use to obtain a serum sample.

3. A kit for detecting HGFAC protein expression in body fluids according to claim 1, characterized in that: Methods for collecting urine samples from body fluid samples include: The patient has not urinated for at least 2 hours. Use a sterile urine collection container to collect 40-50 ml of urine when the patient urinates for the first time in the morning, and immediately put it in a -80℃ refrigerator for later use to obtain a urine sample.

4. A method for constructing an IBD detection model for HGFAC protein expression in body fluids, comprising a HGFAC protein expression detection kit in body fluids as claimed in any one of claims 1 to 3, characterized in that: include: Step 1: Binding the HGFAC protein expression level in a body fluid sample detected by a body fluid HGFAC protein expression detection kit with corresponding clinical data to obtain body fluid data, and establishing an IBD body fluid proteomics database through the body fluid data; the body fluid data includes serum data and urine data; Step 2: Use Jupyter software to import the linear model library, set the random seed and tolerance error to initialize the Logistic regression model; input the serum data and urine data in the IBD body fluid proteomics database into the Logistic regression model for fitting, and output the coefficients and intercepts to obtain the IBD auxiliary diagnosis model; Step 3. If the auxiliary diagnosis result of the IBD auxiliary diagnosis model indicates that there is a risk of IBD, the target body fluid data of the endoscopic score SES-CD before and after conventional treatment that decreases by more than the set threshold is used as the response; the target serum data after conventional treatment is input into the Logistic regression model for fitting, and the coefficients and intercepts are output to obtain the IBD treatment effect monitoring and prognosis prediction model of the serum data; or Input the target urine data after conventional treatment into the Logistic regression model for fitting, and output the coefficients and intercept to obtain the IBD treatment effect monitoring and prognosis prediction model of urine data; or The target serum data and target urine data after conventional treatment were simultaneously input into the Logistic regression model for fitting, and the coefficients and intercepts were output to obtain the IBD treatment effect monitoring and prognosis prediction model combining serum data and urine data; Step 4: Establish ROC curves based on the constructed IBD auxiliary diagnosis model and IBD treatment effect monitoring and prognosis prediction model, respectively. Select the optimal number of random seeds according to the AUC value to obtain the optimal model. Use k-fold cross validation to obtain the optimal model to obtain the best IBD auxiliary diagnosis model and the best IBD treatment effect monitoring and prognosis prediction model.

5. The method for constructing an IBD detection model for HGFAC protein expression in body fluids according to claim 4, characterized in that: In step 1, the body fluid data in the IBD body fluid proteomics database are the body fluid data for which the HGFAC protein expression level in a body fluid detected by the HGFAC protein expression detection kit is less than 3 times the interquartile range and the body fluid data for which the clinical data corresponding to the HGFAC protein expression level is missing are screened out.

6. The method for constructing an IBD detection model for HGFAC protein expression in body fluids according to claim 4, characterized in that: In step 2, the IBD auxiliary diagnosis model is: Y1=a0+a1*x1+a2*y1; Among them, Y1 is the auxiliary diagnosis score of IBD. If Y1≥0, the auxiliary diagnosis result is no risk of IBD. If Y1<0, the auxiliary diagnosis result indicates the risk of IBD. a0 is the intercept of the IBD auxiliary diagnosis model, which is a constant term and represents the predicted value of the IBD auxiliary diagnosis model when all body fluid data are 0; a1 is the regression coefficient of serum data; a2 is the regression coefficient of urine data; x1 is serum data; y1 is urine data.

7. The method for constructing an IBD detection model for HGFAC protein expression in body fluids according to claim 4, characterized in that: In step 3, the IBD treatment effect monitoring and prognosis prediction model of serum data is: Y2=b0+b1*x2; Among them, Y2 is the IBD treatment effect monitoring and prognosis prediction score of serum data. If Y2≥0, it indicates that the patient has a good response to conventional treatment and a good prognosis. If Y2<0, it indicates that the treatment response is not good, the prognosis is poor, and there is a risk of secondary other immune disorders. b0 is the intercept of the IBD treatment effect monitoring and prognosis prediction model of serum data, which is a constant term, indicating the predicted value of the IBD treatment effect monitoring and prognosis prediction model of serum data when the target serum data is 0; b1 is the regression coefficient of the target serum data; x2 is the target serum data.

8. The method for constructing an IBD detection model for HGFAC protein expression in body fluids according to claim 4, characterized in that: In step 3, the IBD treatment effect monitoring and prognosis prediction model of urine data is: Y3=c0+c1*y2; Among them, Y3 is the IBD treatment effect monitoring and prognosis prediction score of urine data. If Y3≥0, it indicates that the patient has a good response to conventional treatment and a good prognosis. If Y3<0, it indicates that the treatment response is not good, the prognosis is poor, and there is a risk of secondary other immune disorders. c0 is the intercept of the IBD treatment effect monitoring and prognosis prediction model of urine data, which is a constant term and represents the predicted value of the IBD treatment effect monitoring and prognosis prediction model of urine data when the target urine data is 0; c1 is the regression coefficient of the target urine data; y2 is the target urine data.

9. The method for constructing an IBD detection model for HGFAC protein expression in body fluids according to claim 4, characterized in that: In step 3, the IBD treatment effect monitoring and prognosis prediction model based on serum and urine data is: Y4=d0+d1*x2+d2*y2; Among them, Y4 is the IBD treatment effect monitoring and prognosis prediction score combined with serum data and urine data. If Y4≥0, it indicates that the patient has a good response to conventional treatment and a good prognosis. If Y4<0, it indicates that the treatment response is not good, the prognosis is poor, and there is a risk of secondary other immune disorders. d0 is the intercept of the IBD treatment effect monitoring and prognosis prediction model based on the combined serum data and urine data, which is a constant term, indicating the predicted value of the IBD treatment effect monitoring and prognosis prediction model based on the combined serum data and urine data when all target body fluid data are 0; d1 is the joint regression coefficient of target serum data; d2 is the joint regression coefficient of the target urine data; x2 is the target serum data; y2 is the target urine data.

10. Use of the method for constructing an IBD detection model for HGFAC protein expression in body fluids as claimed in any one of claims 4 to 9 in auxiliary diagnosis, treatment effect monitoring and prognosis prediction of IBD.

Images