HBV infection degree prediction formula and determination method thereof

Abstract

The invention relates to an HBV infection degree prediction formula and a determination method thereof, and relates to the field of biomedicine. The invention relates to an HBV (Hepatitis B Virus) infection degree prediction formula, which is characterized in that the HBV infection degree prediction formula comprises # imgabs0 #. According to the method, after detection data of healthy people, chronic hepatitis B patients, liver cirrhosis patients and liver cancer patients are collected to obtain multi-dimensional mitochondrial indexes, six strong correlation difference indexes are screened through difference analysis and random forest analysis, a principal component calculation formula is obtained through further PCA dimension reduction analysis, and then an MOPLS (PCA) value is obtained; calculating a calculation formula of the MOPLS value according to the calculated MOPLS (PCA) value, and determining an optimal calculation formula of the MOPLS value through ROC analysis; the method aims at simplifying the calculation process of the MOPLS (PCA) value and obtaining the optimal MOPLS value calculation formula through screening.

Description

Technical Field

[0001] The present application relates to the field of biomedicine, and in particular to a formula for predicting the degree of HBV infection and a method for determining the same. Background Art

[0002] Actively preventing and controlling the prevalence of hepatitis B virus (HBV) infection can reduce the incidence of primary liver cancer, but anti-HBV treatment cannot eradicate the virus. Once immunosuppressed, HBV carriers can experience HBV reactivation and liver dysfunction. Therefore, the detection and follow-up of hepatitis disease warrant greater attention than treatment. Because chronic infection is often asymptomatic until the late clinical stage, simple and reliable monitoring of active hepatitis virus infection is urgently needed.

[0003] In clinical practice, qualitative hepatitis B surface antigen (HBsAg) testing has long been a diagnostic marker for HBV infection. However, these viral detection methods are subject to lags, limitations, and some have significant errors, failing to fully and accurately reflect a patient's true condition. Monitoring disease progression using immune cell mitochondrial markers offers advantages such as high throughput, high speed, low cost, high sensitivity, high precision, a wide linear range, and ease of use. However, research on immune cell mitochondrial function markers remains significantly lacking. Summary of the Invention

[0004] In order to improve the hysteresis and limitations of HBsAg detection, some of which still have large errors and cannot fully and accurately reflect the patient's true condition, and the current research on immune cell mitochondrial function indicators still has significant gaps, this application provides a HBV infection degree prediction formula and a determination method thereof.

[0005] In a first aspect, the present application provides a formula for predicting the degree of HBV infection, using the following technical solution: A HBV infection degree prediction formula is as follows: Any one of the following; Among them, T4Tem.MMP, namely T4Tem.MMP(F), is the percentage of mitochondrial low membrane potential in effector memory helper T cells; T8Tem.PD-1 ＋ .Abs.Count, that is, T8Tem.PD-1 ＋ #(F), absolute counts of PD-1-expressing effector memory killer T cells; CD3 － CD56 ＋ Abs.Count is CD3 － CD56 ＋#(T), absolute NK cell count; T8Tcm.MMP, also known as T8Tcm.MMP(F), is the percentage of low mitochondrial membrane potential in central memory killer T cells; T4Tem.PD-1 ＋ %T4Tem.PD-1 ＋ %(F), percentage of PD-1 expression on effector memory helper T cells; CD3 ＋ .MMP is CD3 ＋ .MMP(T), is the percentage of low mitochondrial membrane potential in T cells.

[0006] By adopting the above technical solutions, the HBV infection degree prediction formula incorporates: T cell mitochondrial membrane potential (MMP): T4Tem.MMP and CD3 ＋ .MMP reflects the degree of mitochondrial functional damage; low mitochondrial membrane potential often indicates abnormal cellular energy metabolism or apoptosis tendency, which is related to the exhaustion of immune cell function caused by HBV infection; T8Tcm.MMP can evaluate the survival ability and antiviral response potential of memory T cells in chronic infection.

[0007] PD-1 expression: T4Tem.PD-1 ＋ % and T8Tem.PD-1 ＋ .Abs.Count reflects the activation level of immune checkpoint molecules; high expression of PD-1 is usually associated with the "exhaustion" state of T cells, indicating immunosuppression caused by persistent HBV infection.

[0008] Absolute NK cell count: NK cells are key effector cells of antiviral natural immunity, and changes in their number can directly reflect the intensity of the body's innate immune response to HBV.

[0009] By incorporating the above-mentioned mitochondrial detection indicators, the HBV infection degree prediction formula can more comprehensively characterize the interaction between the host immune status and viral infection through multi-dimensional integration, avoiding the one-sidedness of a single indicator.

[0010] This application incorporates three core indicators: mitochondrial membrane potential (MMP), PD-1 expression, and NK cell number, to characterize the host immune status from three dimensions: energy metabolism damage, immune cell exhaustion, and inherent immune response intensity, thus avoiding the one-sidedness of a single indicator.

[0011] Preferably, the HBV infection degree prediction formula is: .

[0012] In a second aspect, the present application provides a method for determining a formula for predicting the degree of HBV infection, using the following technical solution: A method for determining a formula for predicting the degree of HBV infection comprises the following steps: S1. Select sample population: The sample population includes healthy people, chronic hepatitis B patients, cirrhosis patients and liver cancer patients.

[0013] S2. Collecting sample data: Collecting blood samples from the sample population and performing mitochondrial index determination to obtain mitochondrial index data for healthy subjects, chronic hepatitis B patients, cirrhosis patients, and liver cancer patients; S3. Screening test indicators: The healthy population, chronic hepatitis B patients, cirrhosis patients, and liver cancer patients were grouped in pairs, and a difference analysis of mitochondrial indicators was performed to screen out common difference indicators that were significantly different in at least five groups; The common difference indicators are screened out by random forest analysis and cross-validation error to obtain strong correlation difference indicators; S4. Analyze indicator data: Performing min-max normalization processing on the data of the strong correlation difference index to obtain normalized data of the strong correlation difference index; Performing PCA dimensionality reduction analysis on the standardized data of the strong correlation difference index to obtain the expression of the principal component; The variance percentage corresponding to the principal component is used as the coefficient of each component, and the MPOLS is obtained by adding them together. （PCA） Value calculation formula; The MPOLS value is the mitochondrial parameter value of the lymphocyte subset; S5. Establish the calculation formula: According to MPOLS （PCA） The MPOLS value calculation formula obtained by the MPOLS value calculation formula is determined by the MPOLS value calculation formula composed of strong correlation difference indicators; Substituting the dimensionality reduction data into the calculation, the MPOLS values ​​of healthy people, chronic hepatitis B patients, cirrhosis patients and liver cancer patients were obtained; The data of mitochondrial indicators corresponding to the healthy population, chronic hepatitis B patients, cirrhosis patients and liver cancer patients were grouped in pairs to obtain six groups, and MPOLS was plotted for each group. （PCA） The ROC curves of the MPOLS value calculation formula and the MPOLS value calculation formula in the six groups are calculated, and the AUC values ​​are calculated. When the AUC values ​​of the MPOLS value calculation formula in the six groups are all greater than those of the MPOLS value calculation formula, the AUC values ​​of the MPOLS value calculation formula are calculated. （PCA） When the AUC value of the calculation formula is calculated, the HBV infection degree prediction formula is obtained.

[0014] By adopting the above technical solution, the sample population includes healthy people, chronic hepatitis B patients, cirrhosis patients, and liver cancer patients, covering different stages of HBV infection, ensuring that the model is suitable for cross-stage prediction of infection severity; by conducting differential analysis in pairwise groups (a total of 6 groups), the bias of comparing single disease groups was avoided.

[0015] By combining differential analysis and random forest analysis to screen mitochondrial indicators, strongly correlated differential indicators were filtered out, greatly improving the specificity of the model; at the same time, min-max normalization processing and PCA dimensionality reduction analysis can further optimize data quality and model structure, simplify computational complexity, retain the core features of the data, and enhance the generalization ability of the computational model.

[0016] Through MOPLS （PCA） The ROC curve and AUC value were compared and optimized based on the HBV infection value calculation formula and the MOPLS value calculation formula to ensure that the HBV infection degree prediction formula has accurate prediction efficacy and can effectively distinguish high-risk and low-risk patients with HBV infection.

[0017] Preferably, the number of common difference indicators screened in step S3 is 33, and the 33 common difference indicators are: T4Tn%(F): percentage of naive helper T cells; T8Tem% (F): percentage of effector memory killer T cells; CD8 ＋ PD-1 ＋ %(F): percentage of PD-1 expression on killer T cells; Lymph.MMP%(F): percentage of lymphocyte mitochondrial low membrane potential; T4Tn.MM(F): mitochondrial mass value of naive helper T cells; T4Tn.MMP(F): percentage of mitochondrial low membrane potential in naive helper T cells; T4Tef.MMP(F): percentage of mitochondrial low membrane potential in effector helper T cells; T4Tem.MMP(F): percentage of mitochondrial low membrane potential in effector memory helper T cells; T8Tn.MMP(F): percentage of mitochondrial low membrane potential of naive killer T cells; T8Tcm.MMP(F): percentage of mitochondrial low membrane potential in central memory killer T cells; T8Tef.MMP(F): percentage of mitochondrial low membrane potential in effector killer T cells; T8Tem.MMP(F): percentage of mitochondrial low membrane potential in effector memory killer T cells; T4Tem.PD-1 ＋%(F): percentage of PD-1 expression on effector memory helper T cells; T8Tem.PD-1 ＋ %(F): percentage of PD-1 expression on effector memory killer T cells; CD45 ＋ #(F): Absolute lymphocyte count (TFUN protocol detection value); Lym%(F): percentage of lymphocytes; CD4 ＋ %(F): percentage of helper T cells; CD4 ＋ #(F): absolute count of helper T cells; T8Tef#(F): absolute count of effector killer T cells; T8Tem#(F): absolute count of effector memory killer T cells; T8Tem.PD-1 ＋ #(F): Absolute counts of PD-1 expressing effector memory killer T cells; T4Tef.MMP#(F): absolute count of cells with low mitochondrial membrane potential in effector helper T cells; T4Tem.MMP#(F): absolute count of cells with low mitochondrial membrane potential in effector memory helper T cells; T8Tef.MMP#(F): absolute count of cells with low mitochondrial membrane potential in effector killer T cells; T8Tem.MMP#(F): absolute count of mitochondrial low membrane potential cells in effector memory killer T cells; T8Tem.PD-1 ＋ MMP(F): percentage of mitochondrial low membrane potential in effector memory killer T cells expressing PD-1; CD3 ＋ #(T): absolute T cell count; CD3 － CD56 ＋ #(T): absolute NK cell count; CD3 ＋ MM(T): T cell mitochondrial mass; Lymph.MMP(T): percentage of lymphocyte mitochondrial low membrane potential; CD3 ＋ MMP(T): percentage of low mitochondrial membrane potential in T cells; CD3 ＋ CD8 ＋ MMP(T): percentage of low membrane potential of mitochondria in killer T cells; Lymph#(T): Absolute lymphocyte count (TBNK protocol detection value).

[0018] Preferably, the strongly correlated difference indicators screened in step S3 are: T4Tem.MMP(F), T8Tem.PD-1 ＋ #(F), CD3 － CD56 ＋ #(T)、T8Tcm.MMP(F)、T4Tem.PD-1 ＋ %(F),CD3 ＋ .MMP(T).

[0019] Preferably, the ethylene propylene rubber is ethylene propylene diene monomer (EPDM). There are 6 main component expressions in step S4, and the expressions of the 6 main components are: y 1 =0.50*x 1 +0.55*x 2 +0.05*x 3 +0.20*x 4 +0.36*x 5 +0.53*x 6 ; y 2 =-0.31*x 1 -0.08*x 2 +0.63*x 3 +0.65*x 4 +0.26*x 5 -0.11x 6 ; y 3 =0.17*x 1 +0.17*x 2 +0.69*x 3 -0.29*x 4 -0.61*x 5 +0.13x 6 ; y 4 =-0.16*x 1 +0.09*x 2 +0.33*x3 -0.66*x 4 +0.63*x 5 -0.16x 6 ; y 5 =0.63*x 1 +0.08*x 2 +0.06*x 3 +0.11*x 4 +0.05*x 5 -0.76x 6 ; y 6 =0.45*x 1 -0.81*x 2 +0.15*x 3 -0.05*x 4 +0.17*x 5 +0.30x 6 ; in, x 1 - x 6 They are T4Tem.MMP(F), T8Tcm.MMP(F), and T4Tem.PD-1 ＋ %(F),T8Tem.PD-1 ＋ #(F), CD3 － CD56 ＋ #(T), CD3 ＋ MMP(T) is the value after normalization; The MPOLS in step S4 （PCA） The value calculation formula is: MPOLS (PCA) = 0.4395*y 1 +0.2208*y 2 +0.1427*y 3 +0.0979*y 4 +0.0540*y 5 +0.4509*y 6 .

[0020] By adopting the above technical solution, 6 strongly correlated difference indicators were obtained through screening. The PCA model constructed based on these 5 strongly correlated difference indicators naturally generated 6 principal components, which completely retained the core characteristic information of the original data. Since the number of principal components was moderate (only 6), there was no need to further screen the cumulative variance percentage to simplify the calculation, which ensured the simplicity and accuracy of the formula calculation.

[0021] Preferably, the number of MPOLS value calculation formulas established in step S5 is 3, and the 3 MPOLS value calculation formulas are: MPOLS value calculation formula 1: ; MPOLS value calculation formula 2: ; MPOLS value calculation formula 3: .

[0022] Preferably, between HCs and CHB, MPOLS (PCA) The AUC value of the ROC curve is 0.8474; The AUC value of the ROC curve for MPOLS1 was 0.8696; The AUC value of the ROC curve of MPOLS2 was 0.8704; The AUC value of the ROC curve of MPOLS3 was 0.9156; Between HCs and LCs, MPOLS (PCA) The AUC value of the ROC curve is 0.9146; The AUC value of the ROC curve for MPOLS1 was 0.9860; The AUC value of the ROC curve of MPOLS2 was 0.9864; The AUC value of the ROC curve of MPOLS3 was 0.9768; Between HCs and HCC, MPOLS (PCA) The AUC value of the ROC curve is 0.9400; The AUC value of the ROC curve for MPOLS1 was 0.9918; The AUC value of the ROC curve of MPOLS2 was 0.9919; The AUC value of the ROC curve of MPOLS3 was 0.9852; Between CHB and LC, MPOLS (PCA)The AUC value of the ROC curve is 0.6444; The AUC value of the ROC curve for MPOLS1 was 0.7819; The AUC value of the ROC curve of MPOLS2 was 0.7798; The AUC value of the ROC curve of MPOLS3 was 0.7645; Between CHB and HCC, MPOLS (PCA) The AUC value of the ROC curve is 0.7815; The AUC value of the ROC curve for MPOLS1 was 0.8873; The AUC value of the ROC curve of MPOLS2 was 0.8894; The AUC value of the ROC curve of MPOLS3 was 0.8083; Between LC and HCC, MPOLS (PCA) The AUC value of the ROC curve is 0.7100; The AUC value of the ROC curve for MPOLS1 was 0.7458; The AUC value of the ROC curve of MPOLS2 was 0.7534; The AUC value of the ROC curve of MPOLS3 was 0.5969; Among them, HCs are healthy people; CHB are patients with chronic hepatitis B; LC are patients with cirrhosis; and HCC are patients with liver cancer.

[0023] By adopting the above technical solutions, MPOLS (PCA) , MPOLS1, MPOLS2 and MPOLS3, the AUC values ​​of MOPLS value calculation formula 1-2 (MOPLS1 and MOPLS2) among each group were always greater than the AUC value of the MOPLS value calculation formula, so MOPLS value calculation formula 1-2 was used as the prediction formula for HBV infection degree; a comprehensive comparison of MOPLS value calculation formula 1-2 showed that the AUC value of MOPLS value calculation formula 2 was better, so MOPLS value calculation formula 2 was preferred.

[0024] Preferably, the MPOLS value of healthy people obtained by the MPOLS calculation formula 2 (MOPLS2) is 0.67 (0.63, 0.71); the MPOLS value of chronic hepatitis B patients is 0.58 (0.51, 0.63); the MPOLS value of cirrhosis patients is 0.50 (0.41, 0.54); and the MPOLS value of liver cancer patients is 0.37 (0.28, 0.49).

[0025] Preferably, when the MPOLS value of the unknown population is lower than 4.51, the unknown population and liver cancer patients are strongly correlated.

[0026] By adopting the above technical solution, it can be considered that the unknown population with a MOPLS value above 0.63 is a healthy population; the unknown population with a MOPLS value between 0.54-0.63 is a chronic hepatitis B patient; the unknown population with a MOPLS value between 0.49-0.51 is a cirrhosis patient; the unknown population with a MOPLS value below 0.41 is a liver cancer patient, and the lower the MOPLS value, the deeper the degree of HBV infection (among which the MOPLS value between 0.51-0.54 is the judgment grayscale for chronic hepatitis B patients and cirrhosis patients, and the MOPLS value between 0.41-0.49 is the judgment grayscale for cirrhosis patients and liver cancer patients).

[0027] In summary, this application has the following beneficial effects: 1. This application incorporates three core indicators, namely mitochondrial membrane potential (MMP), PD-1 expression, and NK cell number, to characterize the host immune status from three dimensions: energy metabolism damage, immune cell exhaustion, and inherent immune response strength, thus avoiding the one-sidedness of a single indicator.

[0028] 2. This application combines differential analysis and random forest analysis to screen detection indicators and filter out the main indicators, greatly improving the specificity of the model. At the same time, min-max normalization processing and PCA dimensionality reduction analysis further simplify the computational complexity while retaining the core features of the data and improving the generalization ability of the computational model. 3. This application simplifies the MOPLS by establishing a MOPLS value calculation formula （PCA） The calculation process of the ROC value and the comparative optimization of the ROC curve and AUC value ensure that the HBV infection degree prediction formula has accurate prediction effectiveness and can effectively distinguish between high-risk and low-risk patients for HBV infection. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a random forest analysis curve diagram in an embodiment of the present application.

[0030] Figure 2 For the MOPLS in this application (PCA) ROC curve of the MOPLS value calculated by the value calculation formula and the MOPLS value formula.

[0031] Figure 3 This is a diagram showing the immune region division and critical value for different disease groups of HBV infection in the examples of this application. DETAILED DESCRIPTION

[0032] The present application is further described in detail below with reference to examples and comparative examples.

[0033] A method for determining a formula for predicting the degree of HBV infection comprises the following steps: S1. Select sample population: The sample population included healthy people, chronic hepatitis B patients, cirrhosis patients and liver cancer patients. Blood samples were drawn from the sample population and tested. The disease diagnosis was confirmed through routine indicators and clinical routine indicators, including five hepatitis B items, liver function tests, HBV viral DNA quantitative detection, imaging and liver biopsy pathology tests. After excluding the influence of other diseases other than HBV infection and / or disease data that interfere with and have a significant impact on immune indicators, 112 healthy people, 173 chronic hepatitis B patients, 64 cirrhosis patients and 110 liver cancer patients were obtained.

[0034] S2. Collect sample data: 20 mL of EDTA-anticoagulated peripheral blood was collected from each sample, and 81 mitochondrial indices were measured. The mitochondrial index data of healthy subjects (HCs), chronic hepatitis B patients (CHB), cirrhosis patients (LC), and liver cancer patients (HCC) were obtained. The specific test indicators and test data are shown in Table 1.

[0035] Table 1. Test indicators and test results of the quartile range

[0036] Note: Marks (F) and (T) indicate two different test schemes, namely the abbreviations of T cell mitochondrial function test scheme (TFUN-mito) and TBNK lymphocyte subset mitochondrial test (TBNK-mito), which are used to distinguish the sources of obtaining the same or similar indicators; MM is mitochondrial mass; # is also marked as Abs. Count is the absolute cell count indicator; T, B, and NK cell ratios are the percentages of each group of cells in the lymphocyte group; T4Tn differentiation indicators are the percentages of each group in the T4 cell group; MMP is MMP low Abbreviation for %, MMP low % is the MMP in the target cell population low Signal is expressed as a percentage of the MMP signal in all target cell populations.

[0037] S3. Screening test indicators: Healthy people, chronic hepatitis B patients, cirrhosis patients, and liver cancer patients were grouped in pairs to obtain six groupings: HCs vs HCB, HCs vs LC, HCs vs HCC, HCB vs LC, HCB vs HCC, and LC vs HCC. The 81 mitochondrial indices were analyzed for differences in the six groups using the t-test method in SPSS software. 33 common difference indices that were different in at least five groups were screened out. The screening results are shown in Table 2.

[0038] Table 2 Results of differential analysis of mitochondrial indicators

[0039] Referring to Table 2, the 33 common difference indicators include: Three items showed significant differences among the six groups: T8Tem.PD-1 + %(F): percentage of PD-1 expression on effector memory killer T cells; T8Tem.MMP(F): percentage of mitochondrial low membrane potential in effector memory killer T cells; T8Tem.PD-1 ＋ MMP(F): percentage of mitochondrial low membrane potential in effector memory killer T cells expressing PD-1; Sixteen items showed significant differences in the five groups: HCs vs CHB, HCs vs LC, HCs vs HCC, CHB vs LC, and CHB vs HCC: Lymph.MMP%(F): percentage of lymphocyte mitochondrial low membrane potential; T4Tn.MMP(F): percentage of mitochondrial low membrane potential in naive helper T cells; T4Tef.MMP(F): percentage of mitochondrial low membrane potential in effector helper T cells; T4Tem.MMP(F): percentage of mitochondrial low membrane potential in effector memory helper T cells; T8Tcm.MMP(F): percentage of mitochondrial low membrane potential in central memory killer T cells; T8Tef.MMP(F): percentage of mitochondrial low membrane potential in effector killer T cells; T8Tem.MMP(F): percentage of mitochondrial low membrane potential in effector memory killer T cells; T8Tef#(F): absolute count of effector killer T cells; T4Tef.MMP#(F): absolute count of cells with low mitochondrial membrane potential in effector helper T cells; T4Tem.MMP#(F): absolute count of cells with low mitochondrial membrane potential in effector memory helper T cells; T8Tef.MMP#(F): absolute count of cells with low mitochondrial membrane potential in effector killer T cells; CD3 - CD56 + #(T): absolute NK cell count; CD3 + MM(T): T cell mitochondrial mass; Lymph.MMP(T): percentage of lymphocyte mitochondrial low membrane potential (TBNK protocol detection value); CD3 + MMP(T): percentage of low mitochondrial membrane potential in T cells; CD3 + CD8 + MMP(T): percentage of low membrane potential of mitochondria in killer T cells; Five items showed significant differences in the five groups: HCs vs CHB, HCs vs LC, HCs vs HCC, CHB vs HCC, and LC vs HCC: T4Tn%(F): percentage of naive helper T cells; T4Tn.MM(F): mitochondrial mass value of naive helper T cells; T8Tem% (F): percentage of effector memory killer T cells; T4Tem.PD-1 ＋ %(F): percentage of PD-1 expression on effector memory helper T cells; Lym%(F): percentage of lymphocytes; One item showed significant differences in all five groups: HCs vs CHB, HCs vs LC, HCs vs HCC, CHB vs LC, and LC vs HCC: CD4 ＋ %(F): percentage of helper T cells; One item showed significant differences in all five groups: HCs vs CHB, HCs vs LC, CHB vs LC, CHB vs HCC, and LC vs HCC: T8Tn.MMP(F): percentage of mitochondrial low membrane potential of naive killer T cells; One item showed significant differences in all five groups: HCs vs CHB, HCs vs HCC, CHB vs LC, CHB vs HCC, and LC vs HCC: CD8 ＋ PD-1 ＋ %(F): percentage of PD-1 expression on killer T cells; Six items showed significant differences in the five groups: HCs vs LC, HCs vs HCC, CHB vs LC, CHB vs HCC, and LC vs HCC: CD45 ＋ #(F): Absolute lymphocyte count (TFUN protocol detection value); CD4+#(F): absolute count of helper T cells; T8Tem#(F): absolute count of effector memory killer T cells; T8Tem.PD-1 ＋ #(F): Absolute counts of PD-1 expressing effector memory killer T cells; CD3 ＋ #(T): Absolute counts of PD-1 expressing effector memory killer T cells; Lymph#(T): Absolute lymphocyte count (TBNK protocol detection value).

[0040] The above 33 common difference indicators were subjected to random forest analysis, with the number of decision trees set to 6000, the cross-validation fold to 10, and the variable reduction amplitude to 1.1. The analysis results are shown in Table 3.

[0041] Table 3 Random forest analysis results of co-variance indices

[0042] According to the above random forest analysis results, the random forest curve is obtained (refer to Figure 1 ), refer to Figure 1 The average decrease in accuracy begins to flatten out at n=6, and the benefits of increasing indicators decrease significantly. Therefore, the six strongly correlated difference indicators with the highest contribution rates are selected, which are: T4Tem.MMP(F): percentage of mitochondrial low membrane potential in effector memory helper T cells; T8Tem.PD-1 ＋ #(F): Absolute counts of PD-1 expressing effector memory killer T cells; CD3 ＋ .MMP(T): percentage of T cell mitochondrial low membrane potential; CD3 － CD56＋ #(T): absolute NK cell count; T4Tem.PD-1 ＋ %(F): percentage of PD-1 expression on effector memory helper T cells; T8Tcm.MMP(F): Percentage of mitochondrial low membrane potential in central memory killer T cells.

[0043] S4. Analyze indicator data: The data of the six strongly correlated difference indicators were processed with min-max standardization using SPSS software to obtain standardized data of the strongly correlated difference indicators and eliminate the influence of unit differences.

[0044] The standardized data of the strongly correlated difference indicators were subjected to PCA dimensionality reduction analysis using SPSS, and 6 principal components (PCs) were obtained, as shown in Table 4.

[0045] Table 4 Total variance explained

[0046] Table 5 Composition matrix

[0047] Referring to Table 5, we can see the load values ​​of each strongly correlated difference indicator variable on different principal components from the component matrix table. According to the knowledge of mathematical statistics, the principal component load matrix can be calculated based on the mathematical relationship between the principal component load matrix, the factor load matrix and the eigenvalue, and the coefficient of each strongly correlated difference indicator in the six principal components can be obtained. The calculation formula is:

[0048] Among them, U is the coefficient, A is the load value of the main index on the main component, and λ is the main component eigenvalue. For example, the coefficient of T4Tem.MMP(F) in the expression of PC1 (take two significant figures after the decimal point) is

[0049] The calculation results of the coefficients of the strong correlation difference index in each principal component expression are shown in Table 6.

[0050] Table 6 Composition coefficient table

[0051] Referring to Table 6, according to the component coefficient table of the main components, 6 main components are obtained y 1 -y The expression for 6 is: y 1 =0.50*x 1 +0.55*x2 +0.05*x 3 +0.20*x 4 +0.36*x 5 +0.53*x 6 ； y 2 =-0.31*x 1 -0.08*x 2 +0.63*x 3 +0.65*x 4 +0.26*x 5 -0.11x 6 ； y 3 =0.17*x 1 +0.17*x 2 +0.69*x 3 -0.29*x 4 -0.61*x 5 +0.13x 6 ； y 4 =-0.16*x 1 +0.09*x 2 +0.33*x 3 -0.66*x 4 +0.63*x 5 -0.16x 6 ； y 5 =0.63*x 1 +0.08*x 2 +0.06*x 3 +0.11*x 4 +0.05*x 5 -0.76x 6 ； y 6 =0.45*x 1 -0.81*x 2+0.15*x 3 -0.05*x 4 +0.17*x 5 +0.30x 6 ; in, x 1 - x 6 They are T4Tem.MMP(F), T8Tcm.MMP(F), and T4Tem.PD-1 ＋ %(F),T8Tem.PD-1 ＋ (F), CD3 － CD56 ＋ (T), CD3 ＋ MMP(T) is the value after normalization; The test data of healthy people, chronic hepatitis B patients, cirrhosis patients and liver cancer patients are brought into the principal component y 1 -y In the expression of 6, the principal component is calculated y 1 -y 6, and some calculation results are shown in Table 7.

[0052] Refer to Table 4-6, take the variance percentages corresponding to the six principal components as their respective coefficients, and add them together to get MPOLS （PCA） Value calculation formula: MPOLS (PCA) = 0.4395*y 1 +0.2208*y 2 +0.1427*y 3 +0.0979*y 4 +0.0540*y 5 +0.4509*y 6 .

[0053] The principal components of healthy people, chronic hepatitis B patients, cirrhosis patients and liver cancer patients were analyzed. y 1 -y Substitute the value of 6 into MPOLS （PCA） The MPOLS of the patient is calculated using the value calculation formula. （PCA） Value, MPOLS of healthy people （PCA） The interquartile range of the value was 0.71 (0.62, 0.82), and the MPOLS of patients with chronic hepatitis B was （PCA）The interquartile range of the values ​​was 0.54 (0.47, 0.62), and the MPOLS of patients with cirrhosis was （PCA） The interquartile range of the values ​​was 0.48 (0.44, 0.57), and the MPOLS of patients with liver cancer was （PCA） The interquartile range of the values ​​was 0.38 (0.22, 0.51). Some of the calculation results are shown in Table 7.

[0054] Table 7 Principal components y 1 -y Partial calculation values ​​and MPOLS of 6 （PCA） The value calculation formula is obtained from the partial MPOLS （PCA） value

[0055] S5. Establish the calculation formula: According to the above MOPLS (PCA) Value calculation results, MOPLS (PCA) The values ​​are ranked from high to low as healthy people (HCs), chronic hepatitis B patients (CHB), cirrhosis patients (LC) and liver cancer patients (HCC).

[0056] Refer to Table 8, T4Tem.MMP, T8Tem.PD-1 ＋ .Abs.Count, CD3 ＋ .MMP, CD3 － CD56 ＋ The numerical trends of Abs.Count and T8Tcm.MMP in healthy subjects (HCs), chronic hepatitis B patients (CHB), cirrhosis patients (LC) and hepatocellular carcinoma (HCC) were compared with those in MOPLS. (PCA) The values ​​are consistent, indicating that these strong correlation difference indicators are consistent with MOPLS (PCA) The values ​​showed a positive correlation trend (due to the killer T cells (CD8 ＋ ) The absolute cell count was significantly increased in CHB infection, while HBV infection caused PD-1 ＋ Therefore, the proportion and absolute count of killer T cells and killer T cell PD-1 in CHB group were significantly increased. ＋ The proportion and absolute count of effector memory killer T cells (T8Tem) were higher than those in the control group HCs (Table 1). ＋The absolute count of T cells decreased significantly, resulting in counts significantly lower than those in the control group HCs), so these strongly correlated difference indicators were incorporated into the MOPLS value calculation formula.

[0057] Table 8 Median values ​​of six strongly correlated difference indicators in HCs, CHB, LC, and HCC

[0058] Among them, T4Tem.MMP is T4Tem.MMP(F); T8Tem.PD-1 ＋ .Abs.Count, that is, T8Tem.PD-1 ＋ #(F); CD3 ＋ .MMP is CD3 ＋ .MMP(T); CD3 － CD56 ＋ Abs.Count is CD3 － CD56 ＋ #(T); T8Tcm.MMP is T8Tcm.MMP(F); T4Tem.PD-1 ＋ %T4Tem.PD-1 ＋ %(F).

[0059] In order to reduce the CV of the data within the group and improve the accuracy (CV is calculated as SD / mean, and the CV>0.5 indicator is added to the denominator calculation formula), in the MOPLS value calculation formula, for T8Tem.PD-1 ＋ .Abs.Count, CD3 － CD56 ＋ Abs.Count、T4Tem.PD-1 ＋ % and T8Tcm.MMP indicators are used to calculate lg (indicator); in order to avoid lg (0) errors or lg (0-1) resulting in negative calculation results, lg (indicator) is adjusted to lg (indicator + 1).

[0060] According to the establishment principle of the above formula, the calculation formula 1 of the MOPLS value is established: .

[0061] Calculate T4Tem.PD-1 in Formula 1 for the MOPLS1 value ＋ %Indicators are simplified and deleted to obtain the MOPLS value calculation formula 2: .

[0062] Based on the MOPLS value calculation formula 2, the T8Tcm.MMP indicator with the lowest contribution rate was simplified and deleted. Due to the extremely high CV value (CV=20.37 / 15.13=1.35) in liver cancer patients, the T8Tem.PD-1 was omitted in the MOPLS value calculation formula 3. ＋ The .Abs.Count indicator is squared to reduce T8Tem.PD-1 ＋ Contribution of .Abs.Count indicator (T8Tem.PD-1 + The absolute cell count of the CHB group increased (Abs.Count), which was higher than that of the HCs group. After the above treatment, the MOPLS calculation formula 3 was obtained: .

[0063] Among them, T4Tem.MMP, T8Tem.PD-1 ＋ .Abs.Count, CD3 ＋ .MMP, CD3 － CD56 ＋ Abs.Count, T8Tcm.MMP and T4Tem.PD-1 ＋ % are used in the above-mentioned MOPLS value calculation formulas 1-3 in the form of numerical values ​​without units (MMP is MMP low %).

[0064] The mitochondrial data of healthy people, chronic hepatitis B patients, cirrhosis patients and liver cancer patients were respectively entered into the MOPLS value calculation formulas 1-3, and the MOPLS values ​​of healthy people, chronic hepatitis B patients, cirrhosis patients and liver cancer patients under each MOPLS value calculation formula were calculated. The calculation results are shown in Table 9.

[0065] Table 9 Interquartile ranges of N values ​​for Group II-IVA and Group IVB calculated using different formulas

[0066] Reference Figure 2 , draw MOPLS respectively （PCA） ROC curves of the three MOPLS value calculation formulas for HCs vs CHB, HCs vs LC, HCs vs HCC, CHB vs LC, CHB vs HCC, and LC vs HCC (where Y （PCA） MOPLS （PCA） ), and calculated the AUC value corresponding to the ROC curve. The AUC value results are shown in Table 10.

[0067] Table 10 AUC values ​​corresponding to ROC curves of different groups for each MOPLS value calculation formula

[0068] Refer to Table 10 and compare with MOPLS （PCA） The AUC values ​​of the value calculation formula and the three MOPLS value calculation formulas among each group were found. The AUC values ​​of the MOPLS value calculation formula 1-2 (MOPLS1 and MOPLS2) among each group were always greater than the AUC values ​​of the MOPLS value calculation formula. Therefore, the MOPLS value calculation formula 1-2 was used as the prediction formula for the degree of HBV infection.

[0069] A comprehensive comparison of MOPLS value calculation formulas 1 and 2 shows that the AUC value of MOPLS value calculation formula 2 is better, so MOPLS value calculation formula 2 is preferred.

[0070] The MOPLS value calculation formula 2 shows that the MOPLS value range of healthy people is 0.63-0.71, the MOPLS value of chronic hepatitis B patients is 0.51-0.63, the MOPLS value of cirrhosis patients is 0.41-0.54, and the MOPLS value of liver cancer patients is 0.28-0.49.

[0071] Reference Figure 3 It can be considered that the unknown population with a MOPLS value above 0.63 is a healthy population; the unknown population with a MOPLS value between 0.54-0.63 is a chronic hepatitis B patient; the unknown population with a MOPLS value between 0.49-0.51 is a cirrhosis patient; the unknown population with a MOPLS value below 0.41 is a liver cancer patient, and the lower the MOPLS value, the deeper the degree of HBV infection (among which the MOPLS value between 0.51-0.54 is the judgment grayscale for chronic hepatitis B patients and cirrhosis patients, and the MOPLS value between 0.41-0.49 is the judgment grayscale for cirrhosis patients and liver cancer patients).

[0072] This specific embodiment is merely an explanation of the present application and is not a limitation of the present application. After reading this specification, those skilled in the art may make non-creative modifications to the present embodiment as needed, but as long as they are within the scope of the claims of the present application, they are protected by the patent law.

Claims

1. A formula for predicting the degree of HBV infection, characterized in that: The HBV infection degree prediction formula is: Any one of the following; Among them, T4Tem.MMP, namely T4Tem.MMP(F), is the percentage of mitochondrial low membrane potential in effector memory helper T cells; T8Tem.PD-1 ＋ .Abs.Count, that is, T8Tem.PD-1 ＋ #(F), absolute counts of PD-1-expressing effector memory killer T cells; CD3 － CD56 ＋ Abs.Count is CD3 － CD56 ＋ #(T), absolute NK cell count; T8Tcm.MMP, also known as T8Tcm.MMP(F), is the percentage of low mitochondrial membrane potential in central memory killer T cells; T4Tem.PD-1 ＋ %T4Tem.PD-1 ＋ %(F), percentage of PD-1 expression on effector memory helper T cells; CD3 ＋ .MMP is CD3 ＋ .MMP(T), is the percentage of low mitochondrial membrane potential in T cells.

2. The HBV infection degree prediction formula according to claim 1, characterized in that: The HBV infection degree prediction formula is:

3. The method for determining the HBV infection degree prediction formula according to any one of claims 1 to 2, characterized in that: The following steps are involved: S1. Select sample population: The sample population includes healthy people, chronic hepatitis B patients, cirrhosis patients, and liver cancer patients; S2. Collecting sample data: Collecting blood samples from the sample population and performing mitochondrial index determination to obtain mitochondrial index data for healthy subjects, chronic hepatitis B patients, cirrhosis patients, and liver cancer patients; S3. Screening test indicators: The healthy population, chronic hepatitis B patients, cirrhosis patients, and liver cancer patients were grouped in pairs, and a difference analysis of mitochondrial indicators was performed to screen out common difference indicators that were significantly different in at least five groups; The common difference indicators are screened out by random forest analysis and cross-validation error to obtain strong correlation difference indicators; S4. Analyze indicator data: Performing min-max normalization processing on the data of the strong correlation difference index to obtain normalized data of the strong correlation difference index; Performing PCA dimensionality reduction analysis on the standardized data of the strong correlation difference index to obtain the expression of the principal component; The variance percentage corresponding to the principal component is used as the coefficient of each component, and the MPOLS is obtained by adding them together. （PCA） Value calculation formula; The MPOLS value is the mitochondrial parameter value of the lymphocyte subset; S5. Establish the calculation formula: According to MPOLS （PCA） The MPOLS value calculation formula obtained by the MPOLS value calculation formula is determined by the MPOLS value calculation formula composed of strong correlation difference indicators; Substituting the dimensionality reduction data into the calculation, the MPOLS values ​​of healthy people, chronic hepatitis B patients, cirrhosis patients and liver cancer patients were obtained; The data of mitochondrial indicators corresponding to the healthy population, chronic hepatitis B patients, cirrhosis patients and liver cancer patients were grouped in pairs to obtain six groups, and MPOLS was plotted for each group. （PCA） The ROC curves of the MPOLS value calculation formula and the MPOLS value calculation formula in the six groups are calculated, and the AUC values ​​are calculated. When the AUC values ​​of the MPOLS value calculation formula in the six groups are all greater than those of the MPOLS value calculation formula, the AUC values ​​of the MPOLS value calculation formula are calculated. （PCA） When the AUC value of the calculation formula is calculated, the HBV infection degree prediction formula is obtained.

4. The method for determining the HBV infection degree prediction formula according to claim 3, characterized in that: The number of common difference indicators screened in step S3 is 33, and the 33 common difference indicators are: T4Tn%(F): percentage of naive helper T cells; T8Tem% (F): percentage of effector memory killer T cells; CD8 ＋ PD-1 ＋ %(F): percentage of PD-1 expression on killer T cells; Lymph.MMP%(F): percentage of lymphocyte mitochondrial low membrane potential; T4Tn.MM(F): mitochondrial mass value of naive helper T cells; T4Tn.MMP(F): percentage of mitochondrial low membrane potential in naive helper T cells; T4Tef.MMP(F): percentage of mitochondrial low membrane potential in effector helper T cells; T4Tem.MMP(F): percentage of mitochondrial low membrane potential in effector memory helper T cells; T8Tn.MMP(F): percentage of mitochondrial low membrane potential of naive killer T cells; T8Tcm.MMP(F): percentage of mitochondrial low membrane potential in central memory killer T cells; T8Tef.MMP(F): percentage of mitochondrial low membrane potential in effector killer T cells; T8Tem.MMP(F): percentage of mitochondrial low membrane potential in effector memory killer T cells; T4Tem.PD-1 ＋ %(F): percentage of PD-1 expression on effector memory helper T cells; T8Tem.PD-1 ＋ %(F): percentage of PD-1 expression on effector memory killer T cells; CD45 ＋ #(F): Absolute lymphocyte count (TFUN protocol detection value); Lym%(F): percentage of lymphocytes; CD4 ＋ %(F): percentage of helper T cells; CD4 ＋ #(F): absolute count of helper T cells; T8Tef#(F): absolute count of effector killer T cells; T8Tem#(F): absolute count of effector memory killer T cells; T8Tem.PD-1 ＋ #(F): Absolute counts of PD-1 expressing effector memory killer T cells; T4Tef.MMP#(F): absolute count of cells with low mitochondrial membrane potential in effector helper T cells; T4Tem.MMP#(F): absolute count of cells with low mitochondrial membrane potential in effector memory helper T cells; T8Tef.MMP#(F): absolute count of cells with low mitochondrial membrane potential in effector killer T cells; T8Tem.MMP#(F): absolute count of mitochondrial low membrane potential cells in effector memory killer T cells; T8Tem.PD-1 ＋ MMP(F): percentage of mitochondrial low membrane potential in effector memory killer T cells expressing PD-1; CD3 ＋ #(T): absolute T cell count; CD3 － CD56 ＋ #(T): absolute NK cell count; CD3 ＋ MM(T): T cell mitochondrial mass; Lymph.MMP(T): percentage of lymphocyte mitochondrial low membrane potential; CD3 ＋ MMP(T): percentage of low mitochondrial membrane potential in T cells; CD3 ＋ CD8 ＋ MMP(T): percentage of low membrane potential of mitochondria in killer T cells; Lymph#(T): Absolute lymphocyte count (TBNK protocol detection value).

5. The method for determining the HBV infection degree prediction formula according to claim 4, characterized in that: The strongly correlated difference indicators screened in step S3 are: T4Tem.MMP(F), T8Tem.PD-1 ＋ #(F), CD3 － CD56 ＋ #(T)、T8Tcm.MMP(F)、T4Tem.PD-1 ＋ %(F),CD3 ＋ .MMP(T).

6. The method for determining the HBV infection degree prediction formula according to claim 4, characterized in that: There are 6 main component expressions in step S4, and the expressions of the 6 main components are: y 1 =0.50*x 1 +0.55*x 2 +0.05*x 3 +0.20*x 4 +0.36*x 5 +0.53*x 6 ； y 2 =-0.31*x 1 -0.08*x 2 +0.63*x 3 +0.65*x 4 +0.26*x 5 -0.11x 6 ； y 3 =0.17*x 1 +0.17*x 2 +0.69*x 3 -0.29*x 4 -0.61*x 5 +0.13x 6 ； y 4 =-0.16*x 1 +0.09*x 2 +0.33*x 3 -0.66*x 4 +0.63*x 5 -0.16x 6 ； y 5 =0.63*x 1 +0.08*x 2 +0.06*x 3 +0.11*x 4 +0.05*x 5 -0.76x 6 ； y 6 =0.45*x 1 -0.81*x 2 +0.15*x 3 -0.05*x 4 +0.17*x 5 +0.30x 6 ； in, x 1 - x 6 They are T4Tem.MMP(F), T8Tcm.MMP(F), and T4Tem.PD-1 ＋ %(F),T8Tem.PD-1 ＋ #(F), CD3 － CD56 ＋ #(T), CD3 ＋ MMP(T) is the value after normalization; The MPOLS in step S4 （PCA） The value calculation formula is: MPOLS (PCA) = 0.4395*y 1 +0.2208*y 2 +0.1427*y 3 +0.0979*y 4 +0.0540*y 5 +0.4509*y 6 。 7. The method for determining the HBV infection degree prediction formula according to claim 6, characterized in that: The number of MPOLS value calculation formulas established in step S5 is 3, and the 3 MPOLS value calculation formulas are: MPOLS value calculation formula 1: ； MPOLS value calculation formula 2: ； MPOLS value calculation formula 3: 。 8. The method for determining the HBV infection degree prediction formula according to claim 7, characterized in that: Between HCs and CHBs, MPOLS (PCA) The AUC value of the ROC curve is 0.8474; The AUC value of the ROC curve for MPOLS1 was 0.8696; The AUC value of the ROC curve of MPOLS2 was 0.8704; The AUC value of the ROC curve of MPOLS3 was 0.9156; Between HCs and LCs, MPOLS (PCA) The AUC value of the ROC curve is 0.9146; The AUC value of the ROC curve for MPOLS1 was 0.9860; The AUC value of the ROC curve of MPOLS2 was 0.9864; The AUC value of the ROC curve of MPOLS3 was 0.9768; Between HCs and HCC, MPOLS (PCA) The AUC value of the ROC curve is 0.9400; The AUC value of the ROC curve for MPOLS1 was 0.9918; The AUC value of the ROC curve of MPOLS2 was 0.9919; The AUC value of the ROC curve of MPOLS3 was 0.9852; Between CHB and LC, MPOLS (PCA) The AUC value of the ROC curve is 0.6444; The AUC value of the ROC curve for MPOLS1 was 0.7819; The AUC value of the ROC curve of MPOLS2 was 0.7798; The AUC value of the ROC curve of MPOLS3 was 0.7645; Between CHB and HCC, MPOLS (PCA) The AUC value of the ROC curve is 0.7815; The AUC value of the ROC curve for MPOLS1 was 0.8873; The AUC value of the ROC curve of MPOLS2 was 0.8894; The AUC value of the ROC curve of MPOLS3 was 0.8083; Between LC and HCC, MPOLS (PCA) The AUC value of the ROC curve is 0.7100; The AUC value of the ROC curve for MPOLS1 was 0.7458; The AUC value of the ROC curve of MPOLS2 was 0.7534; The AUC value of the ROC curve of MPOLS3 was 0.5969; Among them, HCs are healthy people; CHB are patients with chronic hepatitis B; LC are patients with cirrhosis; and HCC are patients with liver cancer.

9. The method for determining the HBV infection degree prediction formula according to claim 8, characterized in that: The MPOLS value of the healthy population obtained by the MPOLS calculation formula 2 (MOPLS2) was 0.67 (0.63, 0.71); the MPOLS value of the chronic hepatitis B patients was 0.58 (0.51, 0.63); the MPOLS value of the cirrhosis patients was 0.50 (0.41, 0.54); and the MPOLS value of the liver cancer patients was 0.37 (0.28, 0.49).

10. The method for determining the HBV infection degree prediction formula according to claim 9, characterized in that: When the MPOLS value of the unknown population was lower than 4.51, there was a strong correlation between the unknown population and liver cancer patients.

Images