A method and system for assessing immune age and status using transcriptomics

Abstract

This invention discloses a method and system for assessing immune age and status using transcriptomics. The invention provides the application of a gene composition highly correlated with age in assessing immune age and status. By utilizing the actual age of the user being tested and the expression levels of the aforementioned gene composition in their peripheral blood mononuclear cell transcriptome, immune age, immune rings, and immune aging index are analyzed to assess immune status. For users whose immune status deviates from the normal range, further classification of their immune status can be determined. This invention employs a machine learning-based approach to effectively extract immune age information from peripheral blood mononuclear cell transcriptome sequencing data, enabling a comprehensive assessment of immune status and providing effective guidance for health status.

Description

Technical Field

[0001] This invention relates to the field of immune function evaluation, and more specifically, to a method and system for assessing immune age and status using transcriptomics. Background Technology

[0002] With the continuous increase in life expectancy, population aging is a result of improved human health, and achieving "healthy aging" is the key to addressing the aging problem.

[0003] Human survival is inextricably linked to the functional immune system. The immune system protects us from infections and malignancies, regulates wound healing, and distinguishes the "self" from surrounding organisms, enabling survival in a competitive environment for space and resources. The innate immune system provides a rapid and effective immune response but lacks discrimination and long-term memory. The acquired immune system, on the other hand, functions through precise antigen recognition, memory formation, and the adaptive proliferation of antigen-specific immune cells. However, during aging, all components of both the innate and adaptive immune systems are affected, manifesting as age-related changes. With increasing age, the immune system gradually loses its ability to respond effectively to pathogens and cancer cells; this decline in immune function with age is known as immunosenescence. The aging of the immune system is one of the causes of illness and death in the elderly. Simultaneously, the aging of immune cells accelerates overall aging.

[0004] Peripheral blood mononuclear cells (PBMCs) are a mixed population of lymphocytes (T cells, B cells, and NK cells), dendritic cells, and monocytes, and are a key component of innate and adaptive immunity. Current techniques for assessing immune status and immune age primarily focus on detecting the proportion of immune cells, failing to accurately assess immune changes at the genetic level. Existing technologies also cannot comprehensively determine the health status and degree of aging of the immune system of the individual being assessed from multiple perspectives, including genes and TCR / BCR changes.

[0005] In summary, current immune status detection technologies cannot accurately assess the age of the immune system from a deeper genetic perspective, or from multiple dimensions, and therefore cannot accurately evaluate the health status of the immune system. Therefore, it is necessary to develop a novel immune status assessment technology to more comprehensively and accurately evaluate the health and aging of the immune system. Summary of the Invention

[0006] To overcome the shortcomings of existing technologies in assessing immune status and to address the problem that existing technologies cannot accurately assess the health status of the immune system, this invention provides a method and system for assessing immune age and status using transcriptomics. This method comprehensively considers changes in genes and TCR / BCR, providing individuals with a more accurate assessment of their immune health. This, in turn, provides a basis for developing targeted interventions and helps people achieve healthy aging.

[0007] The first objective of this invention is to provide a reagent for detecting gene compositions in the assessment of immune age and status.

[0008] A second objective of this invention is to provide a method for assessing immune status using the above-described gene composition.

[0009] A third objective of this invention is to provide a system for assessing immune status using the above-described gene composition.

[0010] A fourth object of the present invention is to provide a computer-readable storage medium.

[0011] The fifth object of the present invention is to provide a computer device.

[0012] To achieve the above objectives, the present invention is implemented through the following solution:

[0013] Since the transcription level of genes in peripheral blood mononuclear cells changes with age, and peripheral blood mononuclear cells encompass a variety of cells in peripheral blood, such as CD8 cells, CD4 cells, NK cells, and B cells, covering most immune cells, these genes are important key indicators of changes in the immune system with age. If the changes in these genes deviate from the normal level for the current age, it indicates a problem with the immune system, thus allowing for a good assessment of a person's current immune status.

[0014] This invention first acquires sequencing data of peripheral blood mononuclear cell transcriptomes from a cohort of healthy individuals aged 5 to 95. Using an algorithm, it extracts the expression levels of key genes in peripheral blood mononuclear cells that change with age. Then, machine learning techniques are used to regress and correlate these key gene expression levels with actual age, thereby establishing an immune age benchmark cohort based on healthy individuals. Based on this immune age benchmark cohort, the immune age of other individuals (such as disease patients) can be accurately predicted, thus precisely assessing an individual's immune status and establishing a predictive framework for changes in their immune status.

[0015] Peripheral blood mononuclear cell (PBMC) immunoage characteristic genes contain crucial information about changes in the human immune system with age. However, changes in a single immunoage characteristic gene are insufficient to cover the immune changes within a specific age group. Therefore, it is necessary to construct an immunoage characteristic regression model by combining mRNA change information from multiple genes to accurately pinpoint immune changes within that age group. This invention utilizes LASSO regression analysis to compress the data, reducing its dimensionality from high to low, avoiding the "curse of dimensionality," and thus improving the accuracy of predictions and the interpretability of the resulting statistical model. It constructs a penalty function to obtain a relatively accurate model, allowing the coefficients of certain gene data to be compressed to 0. Data whose coefficients are not compressed to 0 indicate low collinearity with other data, enabling the screening of genes in the PBMC transcriptome that show significant linear changes with age.

[0016] The application of a detection reagent of a gene composition in assessing immune age and / or immune status, wherein the gene composition comprises the following genes: FAS, PREX2, TRHDE, GPC4, MFN2, LGALSL, CNRIP1, SEMG1, CD70, ANGPTL8, SLC14A2, SFTPD, KRT7, PKIB, GCM1, TTLL8, ADCYAP1, BOC, DAAM2, CDKN2A, SORCS3, TIMP4, PIP, TBX20, BMERB1, LDHD, C11orf40, STARD6, GALNTL6, TMEM119, TACSTD2, and TCN2. The gene composition contains the following genes: C11orf87, NAP1L2, PAQR9, IFNL3, NOS1AP, RNU4-82P, CYP4F30P, TEX26-AS1, LINC01276, LOC389895, OR52I1, C10orf126, UBE2E2-AS1, LINC01524, LINC00607, LINC01554, CT75, RN7SL862P, PPBPP2, BRD9P2, GYPB, OR56A7P, LOC105372440, DSCAS, MEI4, MIR6776, and / or ARHGAP23. The detection reagent detects the expression level of each gene in the gene composition.

[0017] As actual age increases, the above 59 genes show significant linear changes, and their gene expression levels are highly correlated with age changes, but collinearity is low. They are used as immune age characteristic genes to assess immune status.

[0018] A method for assessing immune status using the above-mentioned gene composition includes the following steps:

[0019] Data collection: The actual age of the sample to be tested and the expression level of the gene composition described in claim 1 in the transcriptome of its peripheral blood mononuclear cells were collected;

[0020] Determine immune age:

[0021] Immune age is obtained by using the actual age collected and an immune age-feature gene regression model; wherein, the immune age-feature gene regression model is a model trained on a gradient boosting tree model using data from healthy individuals as the training set.

[0022] Determining immunohistochemical rings:

[0023] The immune rings of the user to be tested can be calculated using the formula: Immune rings = Actual age - Immune age.

[0024] An immune growth ring ≥ 0 is considered a positive growth ring. The initial value of a positive growth ring is one ring. 0 ≤ immune growth ring ≤ +5 is one positive growth ring; +5 < immune growth ring ≤ +10 is two positive growth rings; +10 < immune growth ring ≤ +15 is three positive growth rings; +15 < immune growth ring ≤ +20 is four positive growth rings; +20 < immune growth ring ≤ +25 is five positive growth rings; and immune growth ring > +25 is six or more positive growth rings.

[0025] If the immune ring is less than 0, it is a reverse ring. The initial value of the reverse ring is the first reverse ring. -5 < immune ring < 0 is recorded as the first reverse ring; -25 < immune ring ≤ -5 is the second reverse ring; immune ring ≤ -25 is the third reverse ring.

[0026] Measurement of immune aging deviation index:

[0027] The expression level of the gene composition described in claim 1 in the transcriptome of peripheral blood mononuclear cells was enriched using the ssGSEA algorithm, and the result obtained is the immune aging index.

[0028] The immune aging deviation index of the test sample is calculated according to the formula: Immune aging deviation index = Immune aging index of the test sample - Immune aging index of healthy person.

[0029] Assessing immune status: The immune status of the test sample is assessed using immunohistomorphology and / or immunosenescence deviation index, with the following specific criteria:

[0030] The criteria for assessing the immune status of a sample using its immunohistographs are as follows:

[0031] If the immune system has one forward or one reverse growth ring, then the immune status is excellent.

[0032] If the immune growth rings are two positive growth rings, then the immune status is excellent;

[0033] If the immune growth rings are three positive growth rings, then the immune status is good;

[0034] If the immune growth cycle has four positive growth rings, then the immune status is a risk warning.

[0035] If the immune tree has five positive rings, then the immune status is a high-risk warning sign.

[0036] If the immune tree has six or more positive rings, or two or three negative rings, then the immune status is abnormal.

[0037] The criteria for assessing the immune status of a sample using the immunosenescence deviation index are as follows:

[0038] An immune aging deviation index of -0.25 to 0.05 indicates an excellent immune status.

[0039] An immune aging deviation index of 0.05–0.15 indicates an excellent immune status.

[0040] An immune aging deviation index of 0.15–0.25 indicates a good immune status.

[0041] An immune aging deviation index of 0.25–0.35 indicates a risk warning for the immune status.

[0042] An immune aging deviation index of 0.35–0.5 indicates a high-risk immune status.

[0043] An immune aging deviation index of <-0.35 or >0.5 indicates an immune status that deviates from normal.

[0044] Preferably, the criteria for assessing the immune status of the sample using immunohistochemical rings and the immunosenescence deviation index are as follows:

[0045] An immune status is considered excellent if the immune rings are one cycle of positive or negative growth and the immune aging deviation index is -0.25 to 0.05.

[0046] If the immune rings are positive and the immune aging deviation index is 0.05 to 0.15, then the immune status is excellent.

[0047] If the immune growth rings are three positive growth rings and the immune aging deviation index is 0.15 to 0.25, then the immune status is good.

[0048] If the immune tree has four positive rings and the immune aging deviation index is 0.25–0.35, then the immune status is at risk.

[0049] If the immune tree has five positive rings and the immune aging deviation index is 0.35-0.50, then the immune status is a high-risk warning.

[0050] If the immune growth rings are six or more positive rings, and the immune aging deviation index is >0.5 or <-0.35, then the immune status is deviating from normal.

[0051] The immune cycle consists of three reverse rings. If the immune aging deviation index is >0.5, then the immune status is deviating from normal.

[0052] If the immune rings are the second in reverse order, and the immune aging deviation index is <-0.35, then the immune status is deviating from normal.

[0053] Preferably, the gradient boosting tree model uses an ensemble method of weak learners to predict the immune age, and the model is described by the following simplified formula:

[0054] y=α1*T1(x)+α2*T2(x)+α3*T3(x);

[0055] Where y represents the predicted immune age; x represents the expression level of the 59 genes in the above-mentioned gene composition; T1(x), T2(x), and T3(x) are the basic decision tree models based on x1, x2, and x3, respectively; α1, α2, and α3 represent the weights of each decision tree model.

[0056] More preferably, XGBoost has a learning rate of 0.1, a maximum tree depth of 10, an attribute sampling ratio of 0.8, and 1000 iterations.

[0057] Preferably, the formula for calculating the immune aging index is: Immune aging index = Σ(gene set weight × ssGSEA score), where the gene set is the immune age characteristic gene, the gene set weight is obtained by training and testing the expression data of the immune age characteristic gene using the ssGSEA algorithm, and the ssGSEA score is the expression intensity of the gene set calculated from the gene expression data of the peripheral blood mononuclear cell transcriptome using the ssGSEA algorithm.

[0058] Specifically, the calculation method for the immune aging index is as follows:

[0059] Based on the genes characteristic of immune age, create an immune aging gene set A, and denote the gene list of gene set A as set GL;

[0060] The ssGSEA algorithm is used to sort the expression levels of all genes and obtain the gene rank. The rank is set as BG. Genes existing in BG are found from GL, and the total number of these genes NC and the sum of the expression levels of these genes SG are counted.

[0061] Calculate the enrichment score (ES) of each gene in BG. For any gene G in the expression profile, if G belongs to set GL, the ES value of G is equal to the expression level of the gene divided by SG; otherwise, the ES value of G is equal to 1 divided by (the total number of genes in BG minus NC).

[0062] The ES value with the largest absolute value is taken as the ES value of the immune aging gene set A, i.e., A.ES value;

[0063] Randomly shuffle the order of gene expression levels and recalculate the ES value of the immune aging gene set A; repeat this process 1000 times to obtain 1000 ES values; calculate the position and probability of A's ES value in the distribution of these 1000 ES values, i.e., the p-value;

[0064] The A.ES value is scaled to between 0 and 1 using the min-max normalization method to obtain the immune aging index.

[0065] While immune rings and immune aging deviation index can accurately classify the immune status of most individuals as they age, there are always a few individuals whose immune status changes cannot be accurately determined. In such cases, we can switch to a fuzzy prediction model of immune age status and construct a binary classification task of immune status, namely, an immune classification model. Instead of predicting specific immune age, immune rings, and immune aging deviation index, it predicts whether the immune status is old or young.

[0066] If someone's actual age is young, but their immune status is predicted to be aging, their immune status may differ from that of a healthy person, potentially indicating a certain level of risk. Conversely, if someone's actual age is old, but their immune status is predicted to be young, their immune status may be relatively good. Immune classification models, used in conjunction with immune rings and the immune aging deviation index, serve as a backup plan when immune age predictions are inaccurate, thus improving the overall accuracy of the prediction model. Because changes in immune status are more significant between young and old than in immune age, the prediction results are more precise.

[0067] More preferably, for test samples whose immune status deviates from normal, the following steps are also included:

[0068] Determine actual immune status:

[0069] The actual immune status of the sample is determined by its actual age. The criteria are: 0 ≤ actual age ≤ 65, then the actual immune status is young; actual age > 65, then the actual immune status is old.

[0070] Assessing predictive immune status:

[0071] An immune status classification model is used to predict whether the immune status is young or old; the immune status classification model is a model trained on a binary classification model using data from healthy individuals as the training set.

[0072] Classification for assessing immune status:

[0073] The classification of immune status is assessed using the actual and predicted immune status of the sample to be tested, based on the following specific criteria:

[0074] If the actual immune status is young and the predicted immune status is young, then the immune status is classified as young.

[0075] If the actual immune status is old and the predicted immune status is old, then the immune status is classified as old.

[0076] If the actual immune status is young and the predicted immune status is old, then the immune status is classified as pre-elderly.

[0077] If the actual immune status is old and the predicted immune status is young, then the immune status is classified as young.

[0078] Specifically, the construction of an immune status classification model includes the following steps:

[0079] Import necessary libraries: XGBoost, pandas, sklearn, etc.; Read training data: gene expression data and classification labels; Split the training data into training and test sets in an 8:2 ratio; Construct an XGBoost binary classification model with the objective function "binary classification: logistic regression" and set random states; Train the constructed XGBoost binary classification model on the training set; Use the trained model to make predictions on the test set and obtain predicted labels; Calculate the accuracy between the predicted labels and the true labels; Save the trained XGBoost binary classification model; Load the saved XGBoost binary classification model; Read new data and use the loaded XGBoost binary classification model to predict the new data; Obtain the predicted labels for the new data.

[0080] The formula for the XGBoost binary classification model is as follows:

[0081] For binary classification problems, the output of the XGBoost model is a probability value, expressed as:

[0082]

[0083] Where x is the feature vector, y is the classification label (takes a value of 0 or 1), K is the number of decision trees, and f_k(x) is the prediction result of the k-th decision tree;

[0084] For binary classification problems, XGBoost uses log loss as the objective function:

[0085]

[0086] Where y is the true label and y_hat is the predicted probability;

[0087] XGBoost introduces a regularization term to prevent overfitting. The regularization term includes the complexity of the tree and the sum of squares of the output values ​​of the leaf nodes:

[0088]

[0089] Where T is the number of leaf nodes in the tree, γ is a hyperparameter controlling the complexity of the tree, and w j λ is the output value of the j-th leaf node, and λ is a hyperparameter that controls the sum of squares of the output values;

[0090] Combining all the above parts, the objective function of XGBoost is:

[0091]

[0092] In each iteration, XGBoost attempts to minimize the objective function to improve model performance.

[0093] The binary classification model is described using the following simplified formula:

[0094] P(y)=softmax(α1*T1(x)+α2*T2(x)+α3*T3(x));

[0095] Where P(y) represents the predicted probability of the immune age category; y is the immune age category; x is the expression level of the 59 input genes; T1(x), T2(x), and T3(x) are the basic decision tree models based on x1, x2, and x3, respectively; α1, α2, and α3 represent the weights of each decision tree model; the softmax function is used to convert the model output into a probability distribution; XGBoost has a learning rate of 0.1, a maximum tree depth of 15, an attribute sampling ratio of 0.8, 1000 iterations, a minimum subtree weight of 6, and 2 categories.

[0096] If the immune status of the sample to be tested is determined to be good, excellent, or superb, then relevant data on its actual age and peripheral blood mononuclear cell transcriptome can be collected to update and optimize the cohort of immune age characteristic genes.

[0097] More preferably, the above method further includes:

[0098] For test samples whose immune status is determined to be excellent, good, or good, the expression levels of immune age characteristic genes are updated and / or optimized by using their actual age and the expression levels of the above gene combination in the transcriptome of their peripheral blood mononuclear cells.

[0099] Preferably, the expression level of the transcriptome is TPM.

[0100] The system for assessing immune status using the above-mentioned gene composition includes: an information acquisition module, a storage module, an information analysis module, a judgment module, and an output module;

[0101] The information acquisition module is used to collect the characteristic parameters of the user to be tested, including the actual age of the user to be tested and the expression level of the above gene composition in the transcriptome of its peripheral blood mononuclear cells.

[0102] The storage module is used to store the feature parameters in the information acquisition module, the commands in the information analysis module, and the threshold for determining the immune status of the sample to be tested.

[0103] The information analysis module uses the feature parameters stored in the storage module to analyze and obtain the immune age, immune rings, and / or immune aging deviation index of the user to be tested. The specific analysis method is as follows:

[0104] The immune age is obtained from the characteristic parameters and the immune age-characteristic gene regression model; wherein, the immune age-characteristic gene regression model is a model trained using data from healthy individuals as a training set based on a gradient boosting tree model;

[0105] The immune rings are calculated from the actual age and immune age according to the formula: Immune rings = Actual age - Immune age; where 0 ≤ immune rings ≤ +5 is the first positive ring, +5 < immune rings ≤ +10 is the second positive ring, +10 < immune rings ≤ +15 is the third positive ring, +15 < immune rings ≤ +20 is the fourth positive ring, +20 < immune rings ≤ +25 is the fifth positive ring, immune rings > +25 is the sixth positive ring or more, immune rings < 0 is a reverse ring, -5 < immune rings < 0 is the first reverse ring, -10 < immune rings ≤ -5 is the second reverse ring, and immune rings ≤ -10 is the third reverse ring.

[0106] The immune aging deviation index is calculated according to the formula: immune aging deviation index = immune aging index of the sample to be tested - immune aging index of the healthy person; wherein, the immune aging index is obtained by enrichment analysis of the expression level of the above gene composition in the transcriptome of its peripheral blood mononuclear cells using the ssGSEA algorithm;

[0107] The judgment module is used to compare the immune rings and / or immune aging deviation index obtained by the information analysis module with the threshold in the storage module to determine the immune status of the sample to be tested. The specific criteria are as follows:

[0108] The criteria for assessing the immune status of a sample using its immunohistographs are as follows:

[0109] If the immune system has one forward or one reverse growth ring, then the immune status is excellent.

[0110] If the immune growth rings are two positive growth rings, then the immune status is excellent;

[0111] If the immune growth rings are three positive growth rings, then the immune status is good;

[0112] If the immune growth cycle has four positive growth rings, then the immune status is a risk warning.

[0113] If the immune tree has five positive rings, then the immune status is a high-risk warning sign.

[0114] If the immune tree has six or more positive rings, or two or three negative rings, then the immune status is abnormal.

[0115] The criteria for assessing the immune status of a sample using the immunosenescence deviation index are as follows:

[0116] An immune aging deviation index of -0.25 to 0.05 indicates an excellent immune status.

[0117] An immune aging deviation index of 0.05–0.15 indicates an excellent immune status.

[0118] An immune aging deviation index of 0.15–0.25 indicates a good immune status.

[0119] An immune aging deviation index of 0.25–0.35 indicates a risk warning for the immune status.

[0120] An immune aging deviation index of 0.35–0.5 indicates a high-risk immune status.

[0121] An immune aging deviation index of <-0.35 or >0.5 indicates an immune status that deviates from the normal range.

[0122] The output module is used to output the results obtained by the information analysis module and the judgment module, including: immune rings, immune aging deviation index and / or immune status.

[0123] Preferably, the judgment module is further used to compare the immune rings and immune aging deviation index obtained by the information analysis module with the threshold in the storage module to determine the immune status of the sample to be tested, and the specific criteria are as follows:

[0124] An immune status is considered excellent if the immune rings are one cycle of positive or negative growth and the immune aging deviation index is -0.25 to 0.05.

[0125] If the immune rings are positive and the immune aging deviation index is 0.05 to 0.15, then the immune status is excellent.

[0126] If the immune growth rings are three positive growth rings and the immune aging deviation index is 0.15 to 0.25, then the immune status is good.

[0127] If the immune tree has four positive rings and the immune aging deviation index is 0.25–0.35, then the immune status is at risk.

[0128] If the immune tree has five positive rings and the immune aging deviation index is 0.35-0.50, then the immune status is a high-risk warning.

[0129] If the immune growth rings are six or more positive rings, and the immune aging deviation index is >0.5 or <-0.35, then the immune status is deviating from normal.

[0130] The immune cycle consists of three reverse rings. If the immune aging deviation index is >0.5, then the immune status is deviating from normal.

[0131] If the immune rings are the second in reverse order, and the immune aging deviation index is <-0.35, then the immune status is deviating from normal.

[0132] The output module is used to output the results obtained by the information analysis module and the judgment module, including: immune rings, immune aging deviation index and immune status.

[0133] Preferably, the information analysis module is further used to classify the immune status of the test samples whose immune status is determined to be deviating from the normal in the judgment module, based on their actual immune status and the results of the predicted immune status.

[0134] The criteria for determining the actual immune status are as follows: if 0 ≤ actual age ≤ 65, the actual immune status is young; if actual age > 65, the actual immune status is old.

[0135] An immune status classification model is used to predict whether the immune status is young or old; the immune status classification model is a model trained on a binary classification model using data from healthy individuals as the training set.

[0136] The classification of immune status is assessed using the actual and predicted immune status of the sample to be tested, based on the following specific criteria:

[0137] If the actual immune status is young and the predicted immune status is young, then the immune status is classified as young.

[0138] If the actual immune status is old and the predicted immune status is old, then the immune status is classified as old.

[0139] If the actual immune status is young and the predicted immune status is old, then the immune status is classified as pre-elderly.

[0140] If the actual immune status is old and the predicted immune status is young, then the immune status is classified as young.

[0141] The output module is used to output the results obtained by the information analysis module and the judgment module, including: immune rings, immune aging deviation index, immune status, and classification of immune status.

[0142] Preferably, the output module is used to output the results obtained by the information analysis module and the judgment module, and further includes: immune age and / or immune aging index.

[0143] More preferably, the output module is used to output the results obtained by the information analysis module and the judgment module, including: actual immune status and / or predicted immune status.

[0144] Preferably, the output module is also used to output feature parameters, commands and / or thresholds stored in the storage module.

[0145] Preferably, the system further includes a data update module, used to collect the actual age and gene expression data of peripheral blood mononuclear cell transcriptome of the users to be tested whose immune status is determined to be good, excellent or excellent in the judgment module, and submit them to the information analysis module to update and / or optimize the immune age characteristic genes.

[0146] The system for assessing immune status provided by this invention assesses immune status by collecting sample information, analyzing characteristic parameters, using indicators such as immune rings, immune aging deviation index, and immune status classification, and finally outputs various information.

[0147] A computer-readable storage medium storing a computer program programmed or configured to perform any of the above methods or any of the above systems.

[0148] A computer device includes a memory and a processor, wherein the memory stores a computer program executable on the processor;

[0149] When the computer program is executed by the processor, it implements the operation of any of the above methods or any of the above systems.

[0150] Compared with the prior art, the present invention has the following beneficial effects:

[0151] This invention provides a machine learning-based approach that can effectively obtain immune age information from peripheral blood mononuclear cell transcriptome sequencing data. The immune ring model encompasses important information on gene changes in the transcriptome, enabling a comprehensive assessment of immune status and providing effective guidance for health status. Attached Figure Description

[0152] Figure 1 A flowchart for assessing immune status using transcriptomics.

[0153] Figure 2 The age distribution of 100 healthy individuals aged 5 to 95.

[0154] Figure 3 This is a graph showing the relationship between immune age-related genes and actual age. Basically, all genes change with age, and most genes are closely related to immunity.

[0155] Figure 4 This is a schematic diagram of a tree-based model for predicting immune age.

[0156] Figure 5 This is a flowchart of the immune aging index calculated using the ssGSEA algorithm based on immune characteristic genes.

[0157] Figure 6 An assessment report of the immune status of the user to be tested. Detailed Implementation

[0158] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. These embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Unless otherwise specified, the experimental methods used in the following embodiments are conventional methods; the materials and reagents used, unless otherwise specified, are commercially available.

[0159] Example 1: A method for assessing immune status using transcriptomics

[0160] like Figure 1 As shown, the method for assessing immune status using the transcriptome specifically includes the following steps:

[0161] I. Isolation and sequencing of peripheral blood mononuclear cells (PBMCs)

[0162] (1) Sample Information

[0163] Blood samples were collected from 100 healthy individuals aged 5 to 95. Healthy individuals are defined as those who are physiologically healthy, possessing normal vital signs (such as heart rate, blood pressure, and respiratory rate), a healthy weight and body fat percentage, good lifestyle habits (such as moderate exercise, a healthy diet, and sufficient sleep), and normal reproductive health and immune function.

[0164] like Figure 2 As shown, the age distribution of this group is as follows: 100 people are aged 5-95, 27 people are aged 5-25, 43 people are aged 25-65, and 30 people are aged 65-95. Among them, the largest number of people are aged 25, with a total of 9 people. There are no people aged 34, 75-76, or 88-89.

[0165] (2) Isolation of PBMCs from blood samples

[0166] Blood samples from 100 healthy individuals were transferred to 15 mL centrifuge tubes and diluted with PBS solution at a volume ratio of 1:2. The mixture was then thoroughly combined. An equal volume of lymphocyte separation medium was added to the bottom of each centrifuge tube via a glass Pasteur tube. The tubes were centrifuged at 805 × g for 20 min at room temperature to obtain PBMCs.

[0167] (3) PBMC transcriptome sequencing

[0168] Total RNA was extracted from PBMCs using trizol.

[0169] The total RNA was tested for quality, and if it did not meet the quality control standards (Table 1), the total RNA was extracted again.

[0170] Table 1 RNA Quality Control Standards

[0171]

[0172] Using Illumina The Ultra™ RNA Library Prep Kit further extracts and constructs libraries from the total RNA obtained above (total mass ≥800 ng). The Oligo(dT) magnetic beads in the kit can enrich mRNA containing polyA tails, and the divalent cations can fragment the RNA. The reverse transcriptase then reverse transcribes the fragmented RNA into cDNA. After end repair, A-tailing and sequencing adapter ligation, PCR amplification is performed, and the PCR product is purified to finally obtain the transcriptome library to be sequenced (the final effective concentration of the library is not less than 2 nM).

[0173] PE paired-end sequencing was performed using the Illumina sequencing platform.

[0174] II. Initial Data Acquisition

[0175] (1) Calculate the age of the examinee to obtain the actual age.

[0176] (2) The Illumina sequencer can capture the fluorescence signal of the fluorescently labeled dNTPs during the extension of the sequencing chain. The instrument's built-in software converts the corresponding fluorescence signal into sequencing peaks, thereby obtaining the corresponding sequence information, i.e., obtaining the initial data of the transcriptome in BCL format.

[0177] III. Initial Data Processing

[0178] (1) The initial transcriptome data was converted from BCL format to fastq format using the bcl2fastq2 software, which yielded the raw data. This raw data is a fastq format file containing sequencing reads and corresponding sequencing quality information.

[0179] (2) Use FastQC and TrimGalore software to filter the raw data, remove headers, remove reads containing N, and remove reads below Q20, thus obtaining clean data that can be further analyzed.

[0180] (3) Download the reference genome file and gene annotation file (GTF format) of human (GRCh38 / hg38) from the UCSC website.

[0181] The human genome index was constructed using HISAT2 v2.05 software. The cleandata was then aligned with the human (GRCh38 / hg38) reference genome file using HISAT2 v2.05 software. The alignment results were converted into BAM files using samtools software.

[0182] (4) Use featureCounts software to quantify the number of genes in the bam file.

[0183] First, the number of reads mapped to each gene is calculated to obtain the initial raw counts. Then, the raw counts are corrected based on the gene length in the human (GRCh38 / hg38) gene annotation file to eliminate the influence of sequencing depth and gene length on the read count, and the FPKM value (Fragments Per Kilobase of exon model per Million mapped fragments) of each gene is obtained.

[0184] (5) Convert FPKM values ​​to TPM values ​​to obtain gene expression data of the transcriptome.

[0185] IV. Acquisition of Immune Age-Characteristic Genes

[0186] The LASSO algorithm (Least Absolute Shrinkage and Selection Operator) was used to perform regression analysis on gene expression levels and actual age. The mathematical expression of the regression equation is as follows:

[0187] y = β0 + ∑(βj*xj) + ε;

[0188] Where β is the coefficient, x is the gene expression level, y is the actual age, j is the gene, β0 is the intercept, and βj is the coefficient of the j-th gene; for some genes, β can be reduced to 0. ε (epsiilon) is the error term in the regression equation. In practical applications, this error term is usually assumed to be an independent and identically distributed random variable, following a normal distribution with a mean of 0. The error term represents factors that the model cannot explain, i.e., the difference between the actual observed values ​​and the values ​​predicted by the regression equation. These differences may stem from measurement errors, random noise in the data, or other influencing factors not included in the model.

[0189] This regression equation compresses gene expression data. Data whose coefficients can be compressed to 0 (i.e., β = 0) indicates that the gene may not be related to age changes; gene data whose coefficients cannot be compressed to 0 indicate that the collinearity between gene expression data and age data is relatively small.

[0190] In LASSO regression analysis, the β coefficient is obtained by fitting the data and represents the relationship between the independent variable (here, gene expression level) and the dependent variable (actual age). The β coefficient is solved by minimizing the objective function, which consists of two parts: the residual sum of squares (RSS) and an L1 regularization term (penalty term). The mathematical expression of the objective function is as follows:

[0191] minimize(∑(yi-(β0+∑(βj*xij)))^2+λ*∑|βj|)

[0192] Where yi represents the actual age of the i-th sample, xij represents the expression level of the j-th gene in the i-th sample, β0 is the intercept, βj is the coefficient of the j-th gene, and λ is the contraction amount, which controls the strength of the L1 penalty (i.e., |βj|).

[0193] By optimizing the objective function, the β coefficients corresponding to different genes can be obtained. When λ takes an appropriate value, the β coefficients of some genes will be compressed to 0, indicating that these genes may not be related to age changes; while non-zero β coefficients indicate that the corresponding genes are related to age changes. In this way, immune age characteristic genes related to age changes can be screened out.

[0194] Genes that show significant linear changes with actual age are screened out, and then genes that are highly correlated with age changes but have low collinearity are further screened out, thus obtaining immune age characteristic genes.

[0195] This invention collected sequencing data from the peripheral blood mononuclear cell transcriptomes of 100 healthy individuals aged 5–95 years, and screened out 59 immune age characteristic genes related to age-related changes, as follows:

[0196] FAS gene, PREX2 gene, TRHDE gene, GPC4 gene, MFN2 gene, LGALSL gene, CNRIP1 gene, SEMG1 gene, CD70 gene, ANGPTL8 gene, SLC14A2 gene, SFTPD gene, KRT7 gene, PKIB gene, GCM1 gene, TTLL8 gene, ADCYAP1 gene, BOC gene, DAAM2 gene, CDKN2A gene, SORCS3 gene, TIMP4 gene, PIP gene, TBX20 gene, BMERB1 gene, LDHD gene, C11orf40 gene, STARD6 gene, GALNTL6 gene, TMEM119 gene, TACSTD2 gene, TCN2 gene, C11orf87 Genes, NAP1L2 gene, PAQR9 gene, IFNL3 gene, NOS1AP gene, RNU4-82P gene, CYP4F30P gene, TEX26-AS1 gene, LINC01276 gene, LOC389895 gene, OR52I1 gene, C10orf126 gene, UBE2E2-AS1 gene, LINC01524 gene, LINC00607 gene, LINC01554 gene, CT75 gene, RN7SL862P gene, PPBPP2 gene, BRD9P2 gene, GYPB gene, OR56A7P gene, LOC105372440 gene, DSCAS gene, MEI4 gene, MIR6776 gene and / or ARHGAP23 gene.

[0197] like Figure 3 As shown, all 59 genes are positively correlated with age.

[0198] V. Construction of the Immune Age-Characteristic Gene Regression Model and Prediction of Immune Age

[0199] XGBoost (eXtreme Gradient Boosting) is an ensemble learning algorithm based on gradient boosting. Its main idea is to combine multiple weak learners (usually decision trees) to form a powerful predictive model. XGBoost iteratively adds base learners (such as decision trees) to minimize the objective function. In each iteration, the gradient of the loss function is calculated and used to fit the new base learner.

[0200] like Figure 4 As shown, this invention uses the gradient boosting tree model of the XGBoos algorithm to construct a model of the changes in characteristic genes with age (i.e., the immune age-characteristic gene regression model). The specific formula used is as follows:

[0201] (1) Boosted Trees Model

[0202] Objective Function:

[0203] The objective function consists of two parts: a loss function and a regularization term, as shown in the formula:

[0204] Obj(θ) = L(θ) + Ω(θ);

[0205] Where θ represents the model parameters, L(θ) is the loss function, and Ω(θ) is the regularization term.

[0206] In regression problems, the loss function L(θ) is usually the mean squared error, and the formula is:

[0207] L(θ)=∑(y_i-F(x_i))^2where i=1, 2,...,n;

[0208] Regularization terms are used to prevent overfitting and typically consist of two parts: the complexity of the tree and the output values ​​of the leaf nodes.

[0209] Ω(θ)=γT+(1 / 2)λ∑w_j^2,

[0210] Where T represents the number of leaf nodes in the tree, w_j represents the output value of the j-th leaf node, and γ and λ are regularization parameters.

[0211] (2) Gradient Boosting

[0212] For iteration t, the gradient is: The newly added basic learner needs to fit this gradient: f_t(x)=argmin-f∑[g_i^(t)f(x_i)+(1 / 2)h_i^(t)f^2(x_i)+Ω(f)] where h_i^(t) is the second derivative of the loss function.

[0213] Using the formulas above, XGBoost can optimize the loss function and regularization term in each iteration, thereby obtaining a powerful prediction model. The formulas given here represent the basic principles of the XGBoost algorithm; actual implementations also include many optimizations and improvements.

[0214] XGBoot directly fits the optimal model that can classify young to old, and then saves it in computer storage, which is the immune age-feature gene regression model.

[0215] The immune age-characteristic gene regression model is represented as follows:

[0216] F(x) = ∑f_k(x), where k = 1, 2, ..., K;

[0217] Here, F(x) is the final prediction model, f_k(x) is the k-th basic learner (usually a decision tree), and K represents the total number of basic learners.

[0218] In the final model of this invention, an ensemble learning method based on XGBoost is employed to specifically predict immune age for the 59 immune age characteristic genes. The model is described by the following simplified formula:

[0219] y=α1*T1(x)+α2*T2(x)+α3*T3(x)

[0220] Where y represents the predicted immune age; x represents the expression levels of 59 input genes; T1(x), T2(x), and T3(x) are the basic decision tree models based on x1, x2, and x3, respectively; α1, α2, and α3 represent the weights of each decision tree model. XGBoost has a learning rate of 0.1, a maximum tree depth of 10, an attribute sampling ratio of 0.8, and 1000 iterations.

[0221] This invention provides an efficient and accurate method for predicting the immune age of the aforementioned 59 immune age characteristic genes. Embodiments of this invention cover various applications of using XGBoost or other similar ensemble learning methods for immune age prediction of these 59 genes, including but not limited to biomedical research, clinical practice, and drug development. In practical applications, the model can be further optimized, including using more basic decision tree models, considering more split points, and employing other optimization techniques.

[0222] The age predicted by the immune age-feature gene regression model is the immune age.

[0223] VI. Calculation of Immunological Rings

[0224] The difference between actual age and immune age is defined as the immune ring, i.e., immune ring = actual age - immune age.

[0225] Immune growth rings ≥ 0 are defined as positive growth rings, and immune growth rings < 0 are defined as negative growth rings.

[0226] The initial value of positive growth rings is defined as one positive growth ring. Each additional five immune growth rings is counted as one additional ring: 0 ≤ immune growth rings ≤ +5, counted as one positive growth ring; +5 < immune growth rings ≤ +10, counted as two positive growth rings, and so on. When the number of immune growth rings reaches six or more positive growth rings, the immune status needs to be reassessed.

[0227] The initial value for reverse aging rings is defined as one reverse aging ring, i.e., -5 < immune aging ring < 0, which is recorded as one reverse aging ring; -25 < immune aging ring ≤ -5 is recorded as two reverse aging rings; and immune aging ring ≤ -25 is recorded as three reverse aging rings. When the immune aging ring is three reverse aging rings, the immune status needs to be reassessed.

[0228] VII. Calculation of the Immune Aging Index

[0229] like Figure 5 As shown, this invention uses the ssGSEA (single-sample gene set enrichment analysis) algorithm to perform enrichment analysis on 59 immune age characteristic genes to obtain enrichment scores, which are the immune aging index, and can represent the degree of aging of an individual's immune system.

[0230] The specific steps to obtain the immune aging index are as follows:

[0231] 1. Based on the genes characteristic of immune age, create an immune aging gene set A, and denote the gene list of gene set A as set GL.

[0232] 2. Using the ssGSEA algorithm, sort the expression levels of all genes and obtain their ranks. Denote the ranks as a set, denoted as BG. Find the genes present in BG from GL and count the total number of these genes (NC) and the sum of their expression levels (SG).

[0233] 3. Calculate the enrichment score (ES) of each gene in BG. For any gene G in the expression profile, if G belongs to set GL, then the ES value of G is equal to the expression level of the gene divided by SG; otherwise, the ES value of G is equal to 1 divided by (the total number of genes in BG minus NC).

[0234] 4. The ES value with the largest absolute value is taken as the ES value of the immune aging gene set A (A.ES value).

[0235] 5. Randomly shuffle the order of gene expression levels and recalculate the ES value (A.ES value) of the immune aging gene set A. Repeat this process 1000 times to obtain 1000 ES values. Calculate the position and probability (p-value) of the A.ES value in the distribution of these 1000 ES values.

[0236] 6. The A.ES value is scaled to between 0 and 1 using the min-max normalization method of ssGSEA to obtain the immune aging index.

[0237] VIII. Immune Status Assessment

[0238] (1) Immune status was assessed based on immune rings, and the assessment criteria are shown in Table 2.

[0239] Table 2. Immunoanthropometric Assessment Criteria

[0240]

[0241] (2) Assessment of the immune aging index

[0242] Step six of this embodiment calculates the ssGSEA enrichment score of the sample to obtain the sample's immune aging index.

[0243] In step six of this embodiment, the ssGSEA enrichment score of healthy individuals is calculated to obtain their immune aging index. Here, "healthy individuals" refers to those who are physiologically healthy, possessing normal vital signs (such as heart rate, blood pressure, and respiratory rate), healthy weight and body fat percentage, good lifestyle habits (such as moderate exercise, a healthy diet, and sufficient sleep), and normal reproductive health and immune function.

[0244] Immunosenescence deviation index = Immunosenescence index of the sample under test - Immunosenescence index of healthy individuals.

[0245] Immune status was assessed based on the immune aging deviation index, and the assessment criteria are shown in Table 3.

[0246] Table 3 Assessment Criteria for the Immunosenescence Deviation Index

[0247]

[0248] (3) The immune status was assessed by comprehensively considering various indicators. The assessment criteria are shown in Table 4.

[0249] Table 4. Assessment Criteria for Immunological Status Based on Comprehensive Indicators

[0250]

[0251] IX. Construction of an Immune Status Classification Model

[0252] When the result of assessing immune status based on immune rings and immune aging deviation index is "deviation from normal", an immune classification model can be used to assess immune status.

[0253] (1) Taking 65 years old as the boundary, the actual immune status of those aged 0 to 65 is young, and the actual immune status of those aged 65 and above is old.

[0254] (2) Using the XGBoost algorithm, an immune classification model is constructed for the above-mentioned immune age characteristic genes. The specific steps are as follows:

[0255] 1. Import the required libraries: XGBoost, pandas, sklearn, etc.;

[0256] 2. Read training data: gene expression data and classification labels;

[0257] 3. Divide the training data into a training set and a test set in an 8:2 ratio;

[0258] 4. Construct an XGBoost binary classification model with the objective function "binary classification: logistic regression" and set random states;

[0259] 5. Train the constructed XGBoost binary classification model on the training set;

[0260] 6. Use the trained model to make predictions on the test set to obtain the predicted labels;

[0261] 7. Calculate the accuracy of the predicted labels compared to the actual labels;

[0262] 8. Save the trained XGBoost binary classification model;

[0263] 9. Load the saved XGBoost binary classification model;

[0264] 10. Read new data and use the loaded model to make predictions on the new data;

[0265] 11. Obtain the predicted labels for the new data.

[0266] (3) Formula for XGBoost binary classification model

[0267] 1. Model Representation:

[0268] For binary classification problems, the output of the XGBoost model is a probability value, expressed as:

[0269]

[0270] Where x is the feature vector, y is the classification label (with a value of 0 or 1), K is the number of decision trees, and f_k(x) is the prediction result of the k-th decision tree.

[0271] 2. Objective function:

[0272] For binary classification problems, XGBoost uses log loss as the objective function:

[0273]

[0274] Where y is the true label and y_hat is the predicted probability.

[0275] 3. Regularization term:

[0276] XGBoost introduces a regularization term to prevent overfitting. The regularization term comprises the complexity of the tree and the sum of the squares of the output values ​​of the leaf nodes:

[0277]

[0278] Where T is the number of leaf nodes in the tree, γ is a hyperparameter controlling the complexity of the tree, and w j λ is the output value of the j-th leaf node, and λ is a hyperparameter that controls the sum of squares of the output values.

[0279] Combining all the above parts, the objective function of XGBoost is:

[0280]

[0281] In each iteration, XGBoost attempts to minimize the objective function to improve model performance.

[0282] This algorithm uses xgboost to build a binary classification model, train and evaluate the model, save and load the model, and make predictions on new data.

[0283] In practice, the XGBoost algorithm does not have a fixed formula because it relies on building multiple decision trees. Each tree is optimized based on the residual gradient in each iteration, and these trees are finally combined to obtain the best prediction result.

[0284] In the final model of this invention, an ensemble learning method based on XGBoost is employed to specifically predict the immune age category for the 59 immune age characteristic genes. The classification model is described by the following simplified formula:

[0285] P(y)=softmax(α1*T1(x)+α2*T2(x)+α3*T3(x));

[0286] Where P(y) represents the predicted probability of the immune age category; y is the immune age category; x is the expression level of the 59 input genes; T1(x), T2(x), and T3(x) are the basic decision tree models based on x1, x2, and x3, respectively; α1, α2, and α3 represent the weights of each decision tree model; and the softmax function is used to convert the model output into a probability distribution. XGBoost has a learning rate of 0.1, a maximum tree depth of 15, an attribute sampling ratio of 0.8, 1000 iterations, a minimum subtree weight of 6, and 2 categories.

[0287] This invention provides an efficient and accurate method for predicting the immune age category of 59 immune age characteristic genes. Embodiments of this invention cover various applications of using XGBoost or other similar ensemble learning methods for immune age category prediction of these 59 genes, including but not limited to biomedical research, clinical practice, and drug development. In practical applications, the model can be further optimized, including using more basic decision tree models, considering more split points, and employing other optimization techniques.

[0288] The fitted immune classification model will be stored in computer storage for model prediction.

[0289] (3) When new gene expression data is input, the immune classification model is loaded. The immune classification model predicts the classification result of immune status as young or old.

[0290] Immune status was assessed based on an immune classification model, and the assessment criteria are shown in Table 5.

[0291] Table 5. Criteria for classifying immune status using the immune classification model.

[0292]

[0293]

[0294] 10. Data Update

[0295] If the subject's immune status is determined to be good, excellent, or superb, then their actual age and peripheral blood mononuclear cell transcriptome data can be collected. The transcriptome data can be analyzed according to the regression equation in step four of this embodiment, "Immune Age Characteristic Gene Acquisition," thereby updating and optimizing the immune age characteristic gene cohort.

[0296] Example 2: A method for assessing immune status using transcriptomics

[0297] I. Experimental Methods

[0298] 1. Sample collection

[0299] Blood was collected from six subjects and transferred to 15 mL centrifuge tubes. The blood was diluted with PBS solution at a volume ratio of 1:2 and mixed thoroughly. An equal volume of lymphocyte separation medium was added to the bottom of the centrifuge tube through a glass Pasteur tube. The tubes were centrifuged at 805 × g for 20 min at room temperature to obtain PBMCs. Cell counts were performed, and cell viability was assessed using trypan blue staining. Total RNA was extracted from the PBMCs using Trizol, and transcriptome sequencing was performed to obtain the PBMC transcriptome sequencing data of the subjects.

[0300] 2. Sample testing

[0301] Following the method in Example 1, the PBMC transcriptome sequencing data of the subjects to be tested were analyzed to obtain immune age, immune rings, immune aging index, and immune aging deviation index. The immune status was assessed according to Tables 3-5. When the immune status of the subjects to be tested was determined to be deviating from normal, the immune status was determined to be young or old through the immune classification model.

[0302] II. Experimental Results

[0303] like Figure 6 As shown, Sample 1 has an actual age of 43, an immune age of 45, one positive immune ring, an immune aging deviation index of 0.02, and an excellent immune status; Sample 2 has an actual age of 55, an immune age of 62, two positive immune rings, an immune aging deviation index of 0.05–0.15, and an excellent immune status; Sample 3 has an actual age of 23, an immune age of 34, three positive immune rings, an immune aging deviation index of 0.15–0.25, and a good immune status; Sample 4 has an actual age of 24. Sample 39 had an immune age of 39, four positive immune rings, an immune aging deviation index of 0.25–0.35, and an immune status of risk warning. Sample 5 had an actual age of 23, an immune age of 45, five or more positive immune rings, an immune aging deviation index of 0.35–0.50, and an immune status of high risk warning. Sample 6 had an actual age of 60, an immune age of 25, two reverse immune rings, an immune aging deviation index of <-0.35, and an immune status of deviating from normal, classified as relatively young.

[0304] Example 3: A system for assessing immune status

[0305] This invention provides a system for assessing immune status. The system can be used to perform the transcriptome-based immune status assessment method provided in any of the above embodiments. The system can be implemented in software or hardware and includes: an information acquisition module, a storage module, an information analysis module, a judgment module, an output module, and a data update module.

[0306] (1) Information collection module

[0307] The data is used to collect characteristic parameters of the user to be tested. The characteristic parameters include the actual age of the user to be tested and the expression levels of 59 immune age characteristic genes in the peripheral blood mononuclear cell transcriptome of the sample to be tested in Example 1.

[0308] (2) Storage module

[0309] It is used to store feature parameters in the information acquisition module, commands in the information analysis module, and thresholds used to determine the immune status of the sample to be tested.

[0310] (3) Information Analysis Module

[0311] This is used to analyze the feature parameters in the storage module to obtain the immune age, immune rings, immune aging index, and immune aging deviation index of the user to be tested.

[0312] The immune age characteristic genes are the 59 immune age characteristic genes in Example 1: FAS gene, PREX2 gene, TRHDE gene, GPC4 gene, MFN2 gene, LGALSL gene, CNRIP1 gene, SEMG1 gene, CD70 gene, ANGPTL8 gene, SLC14A2 gene, SFTPD gene, KRT7 gene, PKIB gene, GCM1 gene, TTLL8 gene, ADCYAP1 gene, BOC gene, DAAM2 gene, CDKN2A gene, SORCS3 gene, TIMP4 gene, PIP gene, TBX20 gene, BMERB1 gene, LDHD gene, C11orf40 gene, STARD6 gene, GALNTL6 gene, TMEM119 gene, and TACSTD2 gene. The following genes are included: TCN2, C11orf87, NAP1L2, PAQR9, IFNL3, NOS1AP, RNU4-82P, CYP4F30P, TEX26-AS1, LINC01276, LOC389895, OR52I1, C10orf126, UBE2E2-AS1, LINC01524, LINC00607, LINC01554, CT75, RN7SL862P, PPBPP2, BRD9P2, GYPB, OR56A7P, LOC105372440, DSCAS, MEI4, MIR6776, and ARHGAP23.

[0313] The immune age was predicted by the immune age-feature gene regression model in Example 1.

[0314] The immune rings are calculated from actual age and immune age, and the calculation formula is: Immune rings = Actual age - Immune age; where 0 ≤ immune rings ≤ +5 is the first positive ring, +5 < immune rings ≤ +10 is the second positive ring, +10 < immune rings ≤ +15 is the third positive ring, +15 < immune rings ≤ +20 is the fourth positive ring, +20 < immune rings ≤ +25 is the fifth positive ring, immune rings > +25 is the sixth positive ring or more, immune rings < 0 is a reverse ring, -5 < immune rings < 0 is the first reverse ring, -10 < immune rings ≤ -5 is the second reverse ring, and immune rings ≤ -10 is the third reverse ring.

[0315] The immune aging deviation index is calculated according to the formula: immune aging deviation index = immune aging index of the sample to be tested - immune aging index of the healthy person; wherein, the immune aging index is obtained by enrichment analysis of the expression level of the above gene composition in the transcriptome of its peripheral blood mononuclear cells using the ssGSEA algorithm in Example 1.

[0316] (4) Judgment Module

[0317] This is used to compare the immune rings and / or immune aging deviation index obtained from the information analysis module with the threshold in the storage module to determine the immune status of the sample to be tested. The specific criteria are as follows:

[0318] The criteria for assessing the immune status of a sample using its immunohistographs are as follows:

[0319] If the immune system has one forward or one reverse growth ring, then the immune status is excellent.

[0320] If the immune growth rings are two positive growth rings, then the immune status is excellent;

[0321] If the immune growth rings are three positive growth rings, then the immune status is good;

[0322] If the immune growth cycle has four positive growth rings, then the immune status is a risk warning.

[0323] If the immune tree has five positive rings, then the immune status is a high-risk warning sign.

[0324] If an immune system has six or more positive growth rings, or two or three negative growth rings, then the immune status is considered abnormal.

[0325] The criteria for assessing the immune status of a sample using the immunosenescence deviation index are as follows:

[0326] An immune aging deviation index of -0.25 to 0.05 indicates an excellent immune status.

[0327] An immune aging deviation index of 0.05–0.15 indicates an excellent immune status.

[0328] An immune aging deviation index of 0.15–0.25 indicates a good immune status.

[0329] An immune aging deviation index of 0.25–0.35 indicates a risk warning for the immune status.

[0330] An immune aging deviation index of 0.35–0.5 indicates a high-risk immune status.

[0331] An immune aging deviation index of <-0.35 or >0.5 indicates an immune status that deviates from the normal range.

[0332] For test samples whose immune status is determined to be abnormal, the classification of immune status is determined based on the results of their actual immune status and predicted immune status.

[0333] The criteria for determining the actual immune status are as follows: if 0 ≤ actual age ≤ 65, the actual immune status is young; if actual age > 65, the actual immune status is old.

[0334] The predicted immune status results are obtained by the immune status classification model in Example 1;

[0335] The classification of immune status is determined based on the actual and predicted immune status of the sample to be tested, according to the following specific criteria:

[0336] If the actual immune status is young and the predicted immune status is young, then the immune status is classified as young.

[0337] If the actual immune status is old and the predicted immune status is old, then the immune status is classified as old.

[0338] If the actual immune status is young and the predicted immune status is old, then the immune status is classified as pre-elderly.

[0339] If the actual immune status is old and the predicted immune status is young, then the immune status is classified as young.

[0340] (5) Output module

[0341] The results obtained from the output storage module, information analysis module, and judgment module include: actual age, immune age, immune rings, immune aging deviation index, immune status, actual immune status, predicted immune status, and classification of immune status.

[0342] (6) Data Update Module

[0343] This is used to collect the characteristic parameters of the users to be tested whose immune status is determined to be good, excellent, or outstanding in the judgment module, and submit them to the information analysis module to update and optimize the immune age characteristic genes in the queue.

[0344] In the above embodiments, the modules are divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional module are only for easy differentiation and are not used to limit the scope of protection of the present invention.

[0345] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention. For those skilled in the art, other variations or modifications can be made based on the above description and ideas, and it is neither necessary nor possible to exhaustively describe all implementation methods here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A method for assessing immune age and immune status using transcriptomics, characterized in that, Includes the following steps: The gene composition includes the following genes: FAS, PREX2, TRHDE, GPC4, MFN2, LGALS1, CNRIP1, SEMG1, CD70, ANGPTL8, SLC14A2, SFTPD, KRT7, PKIB, GCM1, TTLL8, ADCYAP1, BOC, DAAM2, CDKN2A, SORCS3, TIMP4, PIP, TBX20, BMERB1, LDHD, C11orf40, STARD6, GALNTL6, TMEM119, TACSTD2, TCN2, and C11o. rf87 gene, NAP1L2 gene, PAQR9 gene, IFNL3 gene, NOS1AP gene, RNU4-82P gene, CYP4F30P gene, TEX26-AS1 gene, LINC01276 gene, LOC389895 gene, OR52I1 gene, C10orf126 gene, UBE2E2-AS1 gene, LINC01524 gene, LINC00607 gene, LINC01554 gene, CT75 gene, RN7SL862P gene, PPBPP2 gene, BRD9P2 gene, GYPB gene, OR56A7P gene, LOC105372440 gene, DSCAS gene, MEI4 gene, MIR6776 gene and ARHGAP23 gene; Data collection: The actual age of the sample to be tested and the expression level of the gene composition in the transcriptome of its peripheral blood mononuclear cells were collected; Immune age determination: Immune age is obtained by using the collected actual age and immune age-feature gene regression model; wherein, the immune age-feature gene regression model is a model trained on a gradient boosting tree model using data from healthy individuals as the training set; Immunochronology: The immunochronology of the user to be tested is calculated according to the formula: Immunochronology = Actual Age - Immunological Age. An immune growth ring ≥ 0 is considered a positive growth ring. The initial value of a positive growth ring is one ring. 0 ≤ immune growth ring ≤ +5 is one positive growth ring; +5 < immune growth ring ≤ +10 is two positive growth rings; +10 < immune growth ring ≤ +15 is three positive growth rings; +15 < immune growth ring ≤ +20 is four positive growth rings; +20 < immune growth ring ≤ +25 is five positive growth rings; and immune growth ring > +25 is six or more positive growth rings. If the immune ring is less than 0, it is a reverse ring. The initial value of the reverse ring is the first reverse ring. -5 < immune ring < 0 is recorded as the first reverse ring; -25 < immune ring ≤ -5 is the second reverse ring; immune ring ≤ -25 is the third reverse ring. Determining the immune aging deviation index: The expression level of the gene composition in the transcriptome of peripheral blood mononuclear cells was enriched using the ssGSEA algorithm, and the result is the immune aging index. The immune aging deviation index of the test sample is calculated according to the formula: Immune aging deviation index = Immune aging index of the test sample - Immune aging index of healthy person. Assessing immune status: The immune status of the test sample is assessed using immunohistomorphology and / or immunosenescence deviation index, with the following specific criteria: The criteria for assessing the immune status of a sample using its immunohistographs are as follows: If the immune system has one forward or one reverse growth ring, then the immune status is excellent. If the immune growth rings are two positive growth rings, then the immune status is excellent; If the immune growth rings are three positive growth rings, then the immune status is good; If the immune growth cycle has four positive growth rings, then the immune status is a risk warning. If the immune tree has five positive rings, then the immune status is a high-risk warning sign. If the immune tree has six or more positive rings, or two or three negative rings, then the immune status is abnormal. The criteria for assessing the immune status of a sample using the immunosenescence deviation index are as follows: An immune aging deviation index of -0.25 to 0.05 indicates an excellent immune status. An immune aging deviation index of 0.05–0.15 indicates an excellent immune status. An immune aging deviation index of 0.15–0.25 indicates a good immune status. An immune aging deviation index of 0.25–0.35 indicates a risk warning for the immune status. An immune aging deviation index of 0.35–0.5 indicates a high-risk immune status. An immune aging deviation index of <-0.35 or >0.5 indicates an immune status that deviates from the normal range.

2. The method according to claim 1, characterized in that, The immune status of the test samples was assessed using immunohistochemical rings and the immunosenescence deviation index, with the following specific criteria: An immune status is considered excellent if the immune rings are one cycle of positive or negative growth and the immune aging deviation index is -0.25 to 0.

05. If the immune rings are positive and the immune aging deviation index is 0.05 to 0.15, then the immune status is excellent. If the immune growth rings are three positive growth rings and the immune aging deviation index is 0.15 to 0.25, then the immune status is good. If the immune tree has four positive rings and the immune aging deviation index is 0.25–0.35, then the immune status is at risk. If the immune tree has five positive rings and the immune aging deviation index is 0.35-0.50, then the immune status is a high-risk warning. If the immune growth rings are six or more positive rings, and the immune aging deviation index is >0.5 or <-0.35, then the immune status is deviating from normal. The immune cycle consists of three reverse rings. If the immune aging deviation index is >0.5, then the immune status is deviating from normal. If the immune rings are the second in reverse order, and the immune aging deviation index is <-0.35, then the immune status is deviating from normal.

3. The method according to claim 1 or 2, characterized in that, For test samples whose immune status deviates from the normal range, the following steps are also included: Determining actual immune status: The actual immune status of the sample is determined by its actual age. The criteria are: 0 ≤ actual age ≤ 65, then the actual immune status is young; actual age > 65, then the actual immune status is old. Determine and predict immune status: The immune status is predicted as young or old using an immune status classification model; the immune status classification model is a model trained on a binary classification model using the actual age of a healthy person and the expression level of the gene composition described in claim 1 in the transcriptome of its peripheral blood mononuclear cells as a training set. Classification of immune status: The classification of immune status is determined by using the actual and predicted immune status of the sample to be tested. The specific criteria are as follows: If the actual immune status is young and the predicted immune status is young, then the immune status is classified as young. If the actual immune status is old and the predicted immune status is old, then the immune status is classified as old. If the actual immune status is young and the predicted immune status is old, then the immune status is classified as pre-elderly. If the actual immune status is old and the predicted immune status is young, then the immune status is classified as young.

4. The method according to claim 1, characterized in that, The expression level of the transcriptome was TPM.

5. A system based on the method of claim 1, characterized in that, include: Information acquisition module, storage module, information analysis module, judgment module, and output module; The information acquisition module is used to collect the characteristic parameters of the user to be tested, including the actual age of the user to be tested and the expression level of the gene composition in the transcriptome of its peripheral blood mononuclear cells. The storage module is used to store the feature parameters in the information acquisition module, the commands in the information analysis module, and the threshold for determining the immune status of the sample to be tested. The information analysis module uses the feature parameters in the storage module to analyze and obtain the immune age, immune rings, and / or immune aging deviation index of the user to be tested. The specific analysis method is as follows: The immune age is obtained from the characteristic parameters and the immune age-characteristic gene regression model; wherein, the immune age-characteristic gene regression model is a model trained using data from healthy individuals as a training set based on a gradient boosting tree model; The immune rings are calculated from the actual age and immune age using the formula: Immune rings = Actual age - Immune age; where 0 ≤ immune rings ≤ +5 is the first positive ring, +5 < immune rings ≤ +10 is the second positive ring, +10 < immune rings ≤ +15 is the third positive ring, +15 < immune rings ≤ +20 is the fourth positive ring, +20 < immune rings ≤ +25 is the fifth positive ring, and immune rings > +25 are the sixth positive ring or more. Immune rings < 0 are reverse rings, -5 < immune rings < 0 is the first reverse ring, -10 < immune rings ≤ -5 is the second reverse ring, and immune rings ≤ -10 is the third reverse ring. The immune aging deviation index is calculated according to the formula: immune aging deviation index = immune aging index of the test sample - immune aging index of the healthy person; wherein, the immune aging index is obtained by enrichment analysis of the expression level of the gene composition in the transcriptome of its peripheral blood mononuclear cells using the ssGSEA algorithm; The judgment module is used to compare the immune rings and / or immune aging deviation index obtained by the information analysis module with the threshold in the storage module to determine the immune status of the sample to be tested. The specific criteria are as follows: The criteria for assessing the immune status of a sample using its immunohistographs are as follows: If the immune system has one forward or one reverse growth ring, then the immune status is excellent. If the immune growth rings are two positive growth rings, then the immune status is excellent; If the immune growth rings are three positive growth rings, then the immune status is good; If the immune growth cycle has four positive growth rings, then the immune status is a risk warning. If the immune tree has five positive rings, then the immune status is a high-risk warning sign. If the immune tree has six or more positive rings, or two or three negative rings, then the immune status is abnormal. The criteria for assessing the immune status of a sample using the immunosenescence deviation index are as follows: An immune aging deviation index of -0.25 to 0.05 indicates an excellent immune status. An immune aging deviation index of 0.05–0.15 indicates an excellent immune status. An immune aging deviation index of 0.15–0.25 indicates a good immune status. An immune aging deviation index of 0.25–0.35 indicates a risk warning for the immune status. An immune aging deviation index of 0.35–0.5 indicates a high-risk immune status. An immune aging deviation index of <-0.35 or >0.5 indicates an immune status that deviates from the normal range. The output module is used to output the results obtained by the information analysis module and the judgment module, including: immune rings, immune aging deviation index and / or immune status.

6. The system according to claim 5, characterized in that, The judgment module is also used to compare the immune rings and immune aging deviation index obtained by the information analysis module with the thresholds in the storage module to evaluate the immune status of the sample to be tested. The specific criteria are as follows: An immune status is considered excellent if the immune rings are one cycle of positive or negative growth and the immune aging deviation index is -0.25 to 0.

05. If the immune rings are positive and the immune aging deviation index is 0.05 to 0.15, then the immune status is excellent. If the immune growth rings are three positive growth rings and the immune aging deviation index is 0.15 to 0.25, then the immune status is good. If the immune tree has four positive rings and the immune aging deviation index is 0.25–0.35, then the immune status is at risk. If the immune tree has five positive rings and the immune aging deviation index is 0.35-0.50, then the immune status is a high-risk warning. If the immune growth rings are six or more positive rings, and the immune aging deviation index is >0.5 or <-0.35, then the immune status is deviating from normal. The immune cycle consists of three reverse rings. If the immune aging deviation index is >0.5, then the immune status is deviating from normal. If the immune rings are the second in reverse order, and the immune aging deviation index is <-0.35, then the immune status is deviating from normal. The output module is used to output the results obtained by the information analysis module and the judgment module, including: immune rings, immune aging deviation index and immune status.

7. The system according to claim 5 or 6, characterized in that, The information analysis module is also used to classify the immune status of the test samples whose immune status is determined to be abnormal in the judgment module based on their actual immune status and predicted immune status. The criteria for determining the actual immune status are: 0 ≤ actual age ≤ 65, then the actual immune status is young; If the actual age is greater than 65, then the actual immune status is old. The immune status classification model is used to predict whether the immune status is young or old. The immune status classification model is a model trained using data from healthy individuals as a training set based on a binary classification model. The classification of immune status is determined by using the actual and predicted immune status of the sample to be tested, and the specific criteria are as follows: If the actual immune status is young and the predicted immune status is young, then the immune status is classified as young. If the actual immune status is old and the predicted immune status is old, then the immune status is classified as old. If the actual immune status is young and the predicted immune status is old, then the immune status is classified as pre-elderly. If the actual immune status is old and the predicted immune status is young, then the immune status is classified as young. The output module is used to output the results obtained by the information analysis module and the judgment module, including: immune rings, immune aging deviation index, immune status, and classification of immune status.

8. An application of the method according to claim 1 in assessing immune age and immune status, characterized in that, The expression levels of each gene in the gene composition were detected using a detection reagent.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that is programmed or configured to perform the method of any one of claims 1 to 4 or the system of any one of claims 5 to 7.

10. A computer device, comprising a memory and a processor, characterized in that, The memory stores computer programs that can be executed on the processor; When the computer program is executed by the processor, it implements the operation of any of the methods described in claims 1 to 4 or any of the systems described in claims 5 to 7.

Images