System and method for predicting dementia or mild cognitive impairment

Abstract

A system for predicting dementia or mild cognitive impairment includes a data collection module for acquiring data on a subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's gender, and a dementia or mild cognitive impairment (MCI) probability calculation module for calculating the data acquired by the data collection module to calculate the probability (p) that the subject will suffer from dementia or MCI. The system can be used as a tool for diagnosing MCI and other dementias, while also being used as a potential drug target for MCI, thereby preventing or delaying the progression of MCI.

Description

[Technical field]

[0001] The present invention relates to a system and method for predicting dementia or mild cognitive impairment. [Background technology]

[0002] Dementia is a major neurocognitive disorder that affects memory, thinking, language, and behavior, all of which interfere with daily life. The 2010 World Alzheimer's Disease Report estimates that due to the aging of the world's population, the economic impact of dementia will be greater than that of cancer, heart disease, and stroke combined. The number of people with dementia worldwide is estimated to be 75 million by 2030 and 135 million by 2050, which will add a huge burden to healthcare and public health systems. Currently, treatments can only control symptoms and slow the progression of dementia, but there is no known cure. At the same time, China's birth rate is declining year by year, and the country is facing a serious situation that dementia can cause: an increasingly smaller number of working-age adults who can provide ongoing care for millions of dementia patients. Early diagnosis and intervention of dementia is therefore crucial.

[0003] The diagnosis of dementia (severe neurocognitive impairment) and mild cognitive impairment (mild neurocognitive impairment / MCI) is based on medical history, examination, assessment of cognitive function, brain imaging, and cerebrospinal fluid (CSF) biomarkers. Alzheimer's disease (AD) is the most common type of dementia, followed by vascular dementia (VaD) and dementia with Lewy bodies (DLB). At present, there is no cure for dementia patients, but early detection makes treatment possible. Cognitive assessment is currently the most convenient and commonly used method to identify these patients. Currently, biomarkers in cerebrospinal fluid are the most reliable indicators for diagnosing AD, which include three core CSF biomarkers, namely amyloid beta (Aβ) protein, total tau protein, and phosphorylated tau protein. However, markers in cerebrospinal fluid are highly invasive and are only useful when clinical cognitive impairment is present. In addition, although several new imaging-based techniques have attracted attention, they are not cost-effective and are not suitable for early screening of Alzheimer's disease. Ultimately, an ideal Alzheimer's disease biomarker should be non-invasive, easy to use, cost-effective, and able to identify early signs of the neurodegenerative process before cognitive abnormalities become clinically evident. Currently, there are no existing non-invasive models that can simultaneously screen for the three major types of dementia and mild cognitive impairment (MCI). [Disclosure of the Invention]

[0004] The objective of the present invention is to provide an effective system that can be used to predict dementia, including the three most common types of dementia, AD, VaD and DLB, and also to predict MCI.

[0005] In summary, the present invention relates to the following:

[0006] 1. A system for predicting dementia or mild cognitive impairment, comprising: a data collection module for obtaining data regarding the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex; A dementia or mild cognitive impairment probability calculation module for calculating the probability (p) that the subject suffers from dementia or mild cognitive impairment by calculating the data acquired by the data collection module; Including, the system. 2. The system according to item 1, wherein the miRNA is one or more selected from hsa-miR-6761-3p, hsa-miR-3173-5p, and hsa-miR-6716-3p, and preferably, the miRNA includes hsa-miR-6761-3p, hsa-miR-3173-5p, and hsa-miR-6716-3p. 3. The subject's apolipoprotein E4 genotype is the typing of apolipoprotein E4 alleles; Item 1. The system according to item 1. 4. The dementia or mild cognitive impairment probability calculation module pre-stores a formula for calculating the dementia probability (p) fitted by logistic regression based on the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's gender data in the existing database; Item 1. The system according to item 1. 5. The formula is the following formula 1: p=1 / [1+e -(i+a*Age+b*性別+c*APOE型+d*hsa-miR-6761-3p+f*hsa-miR-3173-5p+g*hsa-miR-6716-3p) ] (Formula 1) and In the formula, p is the probability of having dementia or mild cognitive impairment, APOE type is the subject's apolipoprotein E4 genotype, and i, a, b, c, d, f, and g are unitless parameters. In the dementia probability calculation module, the values ​​of a, b, c, d, f, and g are obtained based on the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex, and are substituted into Equation 1 for calculation; Item 4. The system according to item 4. 6. i is an arbitrary value selected from -22.77226 to -15.70663, and is preferably -19.23944; a is an arbitrary value selected from 0.1653123 to 0.2425499, and is preferably 0.2039311; d is an arbitrary value selected from 0.8932064 to 1.7053113, and is preferably 1.2992589; f is an arbitrary value selected from -0.722253 to -0.073824, and is preferably -0.398038; g is any value selected from 0.0263939 to 0.5324189, and is preferably 0.2794064; Item 5. The system according to item 5. 7. If the subject is female, the value of b is 0; When the subject is male, b is an arbitrary value selected from -1.144378 to -0.309418, and is preferably -0.726898. Item 5. The system according to item 5. 8. If the subject does not express the APOE4 genotype, the value of c is 0; When the subject's APOE4 genotype is homozygous, the value of c is any value between 1.1429478 and 3.7701044, preferably 2.4565261; When the subject's APOE4 genotype is heterozygous, the value of c is any value between 0.9740697 and 2.038065, preferably 1.5060673. Item 5. The system according to item 5. 9. A grouping module further includes a grouping parameter for default dementia or mild cognitive impairment that is stored in advance, and groups the calculated probability (p) of dementia or mild cognitive impairment of the subject based on the grouping parameter, thereby grouping the risk of the subject suffering from dementia or mild cognitive impairment; Item 1. The system according to item 1. 10. The grouping criteria prestored in the grouping module are as follows: If the calculated probability (p) of the subject having dementia or mild cognitive impairment is less than 10%, the subject is at low risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 10% or more and less than 50%, the subject is at medium risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 50% or more and less than 90%, the subject is at medium to high risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 90% or more, the subject is at high risk of having dementia or mild cognitive impairment. Item 9. The system according to item 9. 11. A data collection step of obtaining data regarding the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex; A dementia or mild cognitive impairment probability calculation step of calculating the probability (p) that the subject suffers from dementia or mild cognitive impairment by calculating the data acquired in the data collection step; A method for predicting dementia or mild cognitive impairment, comprising: 12. The method according to item 11, wherein the miRNA is one or more selected from hsa-miR-6761-3p, hsa-miR-3173-5p, and hsa-miR-6716-3p, and preferably the miRNA includes hsa-miR-6761-3p, hsa-miR-3173-5p, and hsa-miR-6716-3p. 13. The subject's apolipoprotein E4 genotype is a typing of apolipoprotein E4 alleles; Item 12. The method according to item 11. 14. In the dementia or mild cognitive impairment probability calculation step, a formula for calculating the dementia probability (p) fitted by logistic regression based on the data on the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex in the existing database is stored in advance. Item 12. The method according to item 11. 15. The formula is the following formula 1: p=1 / [1+e -(i+a*Age+b*性別+c*APOE型+d*hsa-miR-6761-3p+f*hsa-miR-3173-5p+g*hsa-miR-6716-3p) ] (Formula 1) and In the formula, p is the probability of having dementia or mild cognitive impairment, APOE type is the subject's apolipoprotein E4 genotype, and i, a, b, c, d, f, and g are unitless parameters. In the dementia probability calculation step, the values ​​of a, b, c, d, f, and g are obtained based on the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex, and are substituted into Equation 1 for calculation. Item 15. The method according to item 14. 16.i is an arbitrary value selected from -22.77226 to -15.70663, and is preferably -19.23944; a is an arbitrary value selected from 0.1653123 to 0.2425499, and is preferably 0.2039311; d is an arbitrary value selected from 0.8932064 to 1.7053113, and is preferably 1.2992589; f is an arbitrary value selected from -0.722253 to -0.073824, and is preferably -0.398038; g is any value selected from 0.0263939 to 0.5324189, and is preferably 0.2794064; Item 16. The method according to item 15. 17. If the subject is female, the value of b is 0; When the subject is male, b is an arbitrary value selected from -1.144378 to -0.309418, and is preferably -0.726898. Item 16. The method according to item 15. 18. If the subject does not express the APOE4 genotype, the value of c is 0; When the subject's APOE4 genotype is homozygous, the value of c is any value between 1.1429478 and 3.7701044, preferably 2.4565261; When the subject's APOE4 genotype is heterozygous, the value of c is any value between 0.9740697 and 2.038065, preferably 1.5060673. Item 16. The method according to item 15. 19. A default dementia or mild cognitive impairment grouping parameter is stored in advance, and the calculated dementia or mild cognitive impairment probability (p) of the subject is grouped based on the grouping parameter, thereby grouping the risk of the subject suffering from dementia or mild cognitive impairment. Item 12. The method according to item 11. 20. The grouping criteria prestored in the grouping step are as follows: If the calculated probability (p) of the subject having dementia or mild cognitive impairment is less than 10%, the subject is at low risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 10% or more and less than 50%, the subject is at medium risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 50% or more and less than 90%, the subject is at medium to high risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 90% or more, the subject is at high risk of having dementia or mild cognitive impairment. Item 20. The method according to item 19. [Effects of the invention]

[0007] The present application establishes a mathematical model that predicts whether a subject will suffer from dementia or mild cognitive impairment based on the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's gender. The parameters used in the system established in the present application can be easily and non-invasively detected. The system of the present application is a non-invasive system that can be used to predict the three major types of dementia and is suitable for predicting MCI, and the generalizability of this model is very good, and such a model can be used to perform a universal screening test of the population to identify as many people as possible who are potentially at risk. Among them, the system of the present application is very important for non-predictive MCI, and can be used as a tool to diagnose MCI and other dementias, while it can be used as a potential drug target for MCI, thereby preventing or delaying the progression of MCI. [Brief description of the drawings]

[0008] Various other advantages and merits of the present application will be apparent to those skilled in the art upon reading the detailed description of the preferred embodiments below. The drawings in the specification are only for the purpose of illustrating the preferred embodiments, and are not to be considered as limiting the present application. Apparently, the drawings described below are only some embodiments of the present application, and those skilled in the art can obtain other drawings based on these drawings without creative efforts. In addition, the same parts are given the same reference numerals throughout the drawings. [Figure 1] FIG. 1 shows prediction results of the prediction model in the training set, test set, and validation set of AD. [Diagram 2] FIG. 1 shows the prediction results of the prediction model in the training set, test set, and validation set for VAD, DLB, and MCI. [Diagram 3] FIG. 1 shows the results of grouping the predicted probability of the prediction model and the actual incidence rate in AD data. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] [Details of the invention] Specific embodiments of the present invention are described in more detail below with reference to the accompanying drawings. Although specific embodiments of the present invention are illustrated in the drawings, it should be understood that the present invention can be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0010] It should be noted that in the specification and claims, specific words are used to refer to specific components. Those skilled in the art will understand that different nouns may be used to refer to the same components. In the specification and claims, noun differences are not used as a way to distinguish components, but rather the differences in the functions of components are used as the basis for distinction. Since the terms "including" or "consisting of" referred to throughout the specification and claims are open terms, they should be interpreted as "including but not limited to". The following description is a preferred embodiment for carrying out the present invention, but these descriptions are intended for the general principles of the specification and do not limit the scope of the present invention. The scope of protection of the present invention shall be determined by the appended claims.

[0011] Types of variables: In statistics, variables can be classified into two types: quantitative variables and qualitative variables (also called categorical variables).

[0012] Quantitative variables are variables used to describe the amount or number of things, and can be classified as continuous or discrete. A continuous variable refers to a variable that can take any value within a certain interval, whose value is continuous and can contain decimal points. For example, blood pressure value, blood glucose value, anthropometric height, weight, chest circumference, etc. are continuous variables, whose values ​​can only be obtained by measurement or weighing methods. A discrete variable is a variable whose values ​​can only be natural numbers or integer units. For example, pain scores, number of metastatic lesions, number of eggs retrieved, etc. can only be positive numbers without decimal points, and the values ​​of such variables are usually obtained by counting methods.

[0013] The type of variables is not static, but can be converted between different types of variables according to the needs of the research objective. For example, the amount of hemoglobin (g / L) is originally a numerical variable, but if it is divided into two types, normal hemoglobin and low hemoglobin, the data can be analyzed according to the binary classification, and if it is divided into five levels, severe anemia, moderate anemia, mild anemia, normal, and increased hemoglobin, the data can be analyzed according to the levels. Categorical data can also be quantified. For example, if a patient's nausea response can be expressed as 0, 1, 2, or 3, it can be analyzed as a numerical variable data (quantitative data).

[0014] Logistic regression is a generalized linear regression analysis model that is often used in fields such as data mining, automatic disease diagnosis, and economic forecasting. For example, we investigate the risk factors that cause diseases and predict the probability of disease occurrence based on the risk factors. Take the analysis of gastric cancer status as an example, select two groups of people. One group is the gastric cancer group, and the other group is the non-gastric cancer group. The two groups of people must have different physical signs and lifestyles, etc. Therefore, the dependent variable is the presence or absence of gastric cancer, whose value is "yes" or "no", and the independent variables include many variables, such as age, sex, dietary habits, Helicobacter pylori infection, etc. The independent variables can be either continuous variables or categorical variables. Then, the weights of the independent variables can be obtained through logistic regression analysis, and it is possible to roughly grasp which factors are risk factors for gastric cancer. At the same time, according to this weighting, the probability that a person will suffer from cancer based on the risk factors can be predicted. The dependent variable of logistic regression can be binary or multi-category.

[0015] The model used herein to fit the data through logistic regression is a logistic regression model that penalizes the absolute magnitude of the regression model coefficients based on the value of λ: the larger the penalty, the closer the estimates of weak factors will be to zero, so that only the strongest predictors remain in the model.

[0016] MicroRNA (miRNA) is an endogenous non-coding RNA of about 22 nucleotides that can control gene expression at the post-transcriptional level. Due to its small molecular weight, it can move away from the cell membrane and through the blood circulation. Therefore, miRNA can serve as a powerful tool for non-invasive screening test of disease. Increasing evidence suggests that microRNA plays an important role in various pathological processes throughout the progression of Alzheimer's disease. In this application, the inventor attempts to use miRNA and other basic clinical information to establish a non-invasive model for early identification of dementia, thereby contributing to early intervention of dementia-related symptoms.

[0017] Herein, serum miRNAs were mainly derived from microvesicles-mediated active secretion. They show remarkable long-term stability in the extracellular environment, mainly due to interactions with Argonaute2-miRNA complexes, lipoprotein complexes, or vesicles. Thus, miRNAs as major regulators of gene expression are increasingly recognized as new promising diagnostic biomarker candidates that are non-invasive, inexpensive, and highly sensitive. Differently expressed serum miRNAs, such as miR-31, -93, -143, -146a, -135a, -193b, and -384, have been previously found in AD patients, although the relative same sample amount was used. The applicants of the present application have made significant progress in the data sources studied, using large-scale prospective cohort data to build a single model for predicting various dementia subtypes and mild cognitive impairment.

[0018] The data source team of this application used serum miRNA expression data from 1,601 Japanese people, and established three different models using data from AD, VaD, and DLB and normal control populations, respectively. The AUCs of the AD, VaD, and DLB models were 0.874, 0.867, and 0.870, respectively. The number of miRNAs discovered by the three predictive models was 78 miRNAs, 86 miRNAs, and 110 miRNAs, respectively. The authors of the data source team of this application established three different models for three different types of dementia data, and these models contain many independent variables, which generally means that they have little practical importance in terms of statistics and practical experience, while different models use different predictive variables and no common rules were found for different dementias, which also suggests that the application value of the models is low. In this application, the most common type of dementia, AD dementia, was used in the model, and three major miRNAs were discovered, and the model was verified using internal and external data, suggesting that the stability of the model is improved. Furthermore, the established model is not only effective for AD dementia, but also shows excellent predictive effects in various dementia types and MCI, suggesting that the three miRNAs discovered by the model of this application may discover common patterns in various types of dementia and mild cognitive impairment. Therefore, it has more significant meaning and may be used in the future for diagnosis and screening tests of dementia and mild cognitive impairment, and even as targets for drug development. It has potentially great social and economic importance.

[0019] The greatest risk factor for neurodegenerative diseases is advanced age. APOE epsilon4 (apolipoprotein E4, APOE4) genotype and female gender are two well-known risk factors for age-related AD. Menopause and ovarian estrogen deficiency are speculated to be responsible for the increased incidence of Alzheimer's disease in women aged 65 years and older. Estrogen is the major female hormone and plays important roles such as neuroprotection in both reproductive and non-reproductive systems. Estrogen treatment initiated early in menopause has beneficial neuroprotective effects when neurons are in a healthy state. APOE epsilon4 allele status patterns were also found to be associated with estrogen treatment. Furthermore, subjects with the APOE epsilon4 genotype had a 15-fold increased risk compared to the normal genotype. APOE epsilon4 also contributes to the progression of atherosclerosis and neurodegenerative diseases.

[0020] In order to solve the problems existing in the existing technology, the present application provides a system for predicting dementia or mild cognitive impairment, the system includes a data collection module for acquiring data regarding a subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's gender, and a dementia or mild cognitive impairment probability calculation module for calculating the data acquired by the data collection module to calculate the probability (p) that the subject will suffer from dementia or mild cognitive impairment.

[0021] As used herein, there is no limitation on the data collection module, so long as it can be used to obtain data regarding a subject's miRNA levels, the subject's apolipoprotein E4 (APOE4) genotype, the subject's age, and the subject's gender.

[0022] Specifically, the subject's miRNA levels obtained by the data collection module refer to the abundance of miRNA in serum, which can be detected using existing methods such as sequencing, PCR, or miRNA chip methods.

[0023] If there are multiple miRNAs, the data collection module will obtain the expression level of each miRNA separately. The miRNA levels of subjects can be obtained through existing chip detection.

[0024] In one specific embodiment, the miRNA is one or more selected from hsa-miR-6761-3p, hsa-miR-3173-5p, and hsa-miR-6716-3p.

[0025] In one specific embodiment, the miRNAs include hsa-miR-6761-3p, hsa-miR-3173-5p, and hsa-miR-6716-3p. At this time, the data collection module needs to obtain the levels of three miRNAs, hsa-miR-6761-3p, hsa-miR-3173-5p, and hsa-miR-6716-3p, respectively, in the subject.

[0026] In this application, a large number of miRNAs were screened rudimentarily, but after thorough research by the applicant, it was found that the most significant effect was obtained when one, two or three of the three miRNAs hsa-miR-6761-3p, hsa-miR-3173-5p and hsa-miR-6716-3p were selected and used, and the method and system of this application could be used to predict the three major types of dementia and mild cognitive impairment (MCI). Although there have been studies in the prior art using different miRNAs for prediction, a universal system and method that can be used for early screening tests for Alzheimer's disease has not yet been established. In this application, through extensive experiments, three miRNAs were finally selected from the 2562 miRNAs in the miRNA chip, and the method and system of this application were constructed. The main types of dementia are AD, VaD dementia and DLB dementia. Mild cognitive impairment refers to a disease state between normal aging and dementia. Compared with normal elderly people matched for age and education level, the patients showed mild cognitive decline, but their daily abilities were not significantly affected. The core symptom of mild cognitive impairment is cognitive decline, which, depending on the cause or site of brain damage, may affect one or more of memory, executive function, language, application, visuospatial structural skills, etc., resulting in corresponding clinical symptoms. The prior art lacks methods and systems for comprehensive and universal screening tests for the three major types of Alzheimer's disease and mild cognitive impairment, and the method and system of the present application fills such a gap.

[0027] In the present application, the module for calculating the probability of dementia or mild cognitive impairment calculates the probability (p) that a subject suffers from dementia or mild cognitive impairment by calculating the above data obtained by the data collection module. First, it should be understood that this module has a formula for calculating the probability of dementia (p) that is fitted by logistic regression based on the data on the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex in an existing database. Using such a pre-stored formula, calculation can be performed for any subject.

[0028] In the present invention, the existing database refers to a database composed of obtainable subjects, and there is no limitation on the sample size of the database, but the larger the sample size of the database, the better. For example, the number of subjects may be 100, 200, or 300, preferably 400 or more, and more preferably 500 or more.

[0029] In calculating, this pre-stored formula is a formula that uses data collected by the data collection module regarding the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's gender to calculate the probability that the subject will suffer from dementia.

[0030] Among these, the subject's sex is a binary variable, the subject's apolipoprotein E4 allele status is a ternary variable, the subject's miRNA level and the subject's age are continuous variables.

[0031] Furthermore, the inventors of the present application have constructed a specific formula for predicting the probability (p) that a subject will suffer from dementia. The specific formula is the following Formula 1: p=1 / [1+e -(i+a*Age+b*性別+c*APOE型+d*hsa-miR-6761-3p+f*hsa-miR-3173-5p+g*hsa-miR-6716-3p) ] (Formula 1) and Further, in the above formula 1, p is the probability of dementia or mild cognitive impairment, APOE type is the state of the apolipoprotein E4 gene of the subject, and i, a, b, c, d, f, and g are unitless parameters; In the dementia probability calculation module, the values ​​of a, b, c, d, f, and g are obtained based on the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's gender, and are substituted into Equation 1 for calculation.

[0032] In one particular embodiment, i is an arbitrary value selected from −22.77226 to −15.70663, and is preferably −19.23944; a is an arbitrary value selected from 0.1653123 to 0.2425499, and is preferably 0.2039311; d is an arbitrary value selected from 0.8932064 to 1.7053113, and is preferably 1.2992589; f is an arbitrary value selected from -0.722253 to -0.073824, and is preferably -0.398038; g is an arbitrary value selected from 0.0263939 to 0.5324189, and is preferably 0.2794064.

[0033] In one particular embodiment, the subject's gender is a binary variable, and if the subject is female, the value of b is 0, and if the subject is male, b is any value selected from -1.144378 to -0.309418, preferably -0.726898.

[0034] In one particular embodiment, the subject's apolipoprotein E allele status is a three-valued variable, where if the subject does not express the APOE4 genotype, the value of c is 0, if the subject's APOE4 genotype is homozygous, the value of c is any value between 1.1429478 and 3.7701044, preferably 2.4565261, and if the subject's APOE4 genotype is heterozygous, the value of c is any value between 0.9740697 and 2.038065, preferably 1.5060673.

[0035] Furthermore, the system of the present application may further include a grouping module, in which default grouping parameters for dementia or mild cognitive impairment are stored in advance, and the grouping module groups the calculated probability (p) of dementia or mild cognitive impairment of the subject based on the grouping parameters, thereby grouping the risk of the subject suffering from dementia or mild cognitive impairment.

[0036] The grouping criteria previously stored in the grouping module are as follows. If the calculated probability (p) of the subject having dementia or mild cognitive impairment is less than 10%, the subject is at low risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 10% or more and less than 50%, the subject is at medium risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 50% or more and less than 90%, the subject is at medium to high risk of having dementia or mild cognitive impairment; When the calculated probability (p) of a subject having dementia or mild cognitive impairment is 90% or higher, the subject is at high risk of having dementia or mild cognitive impairment.

[0037] The present application also describes a method for determining whether a subject's miRNA levels are sufficient to determine whether the subject's apolipoprotein E4 allele status is ... A dementia or early cognitive impairment probability calculation step for calculating the probability (p) that the subject will suffer from dementia or early cognitive impairment by calculating the data acquired in the data collection step; The present invention relates to a method for predicting dementia or mild cognitive impairment, comprising:

[0038] As described above, for specific details of the data collection step and the dementia or early cognitive impairment probability calculation step involved in the method for predicting dementia or early cognitive impairment of the present application, such as obtaining data on the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex, and calculating the probability (p) that the subject will suffer from dementia or early cognitive impairment, refer to the above description of each module of the system of the present application. EXAMPLES

[0039] Data used to build the model The model-building data included 1,309 samples from 1,021 AD patients and 288 healthy controls. In addition, 91 VaD cases, 169 DLB cases, and 32 MCI cases were used to evaluate the performance of the AD prediction model in other types of dementia and mild cognitive impairment.

[0040] The serum miRNA chip data of the above subjects and the corresponding age, sex, and APOE allele type were downloaded from GEO (Gene Expression Omnibus) and the accession number is GSE120584.

[0041] According to GSE120584 and related literature (Shigemizu D, Akiyama S, Asanomi Y, Boroevich KA, Sharma A, Tsunoda T, Matsukuma K, Ichikawa M, Sudo H, Takizawa S et al: Risk prediction models for dementia constructed by supervised principal component analysis using miRNA expression data. Commun Biol 2019,2:77.), all 1,601 subjects were over 60 years old, had their APOE4 genotype detected, and underwent the Mini-Mental State Exam (MMSE).

[0042] The diagnosis of all patients and healthy controls was based on medical history, physiological examination, diagnostic tests, neurological examination, neuropsychological examination, and brain imaging by magnetic resonance imaging (MRI) or computed tomography (CT). Neuropsychological examinations included the MMSE, the Japanese version of the Alzheimer's Disease Assessment Scale Cognitive Component, the Wechsler Memory Scale-Revised Logical Memory I and II, the Frontal Appraisal Battery, Raven's Colored Progressive Matrices, and the Geriatric Depression Scale. When necessary, dopamine transporter imaging and metaiodobenzylguanidine myocardial scintigraphy were used to diagnose DLB. Cerebrospinal fluid biomarkers and pathological tests were not used to diagnose dementia.

[0043] The AD cases in this study were probable AD. The diagnosis of AD and MCI was based on the criteria of the National Institute on Aging-Alzheimer's Association workgroup (McKhann GM, Knopman DS, Chertkow H, Hyman BT, Jack CR, Jr., Kawas CH, Klunk WE, Koroshetz WJ, Manly JJ, Mayeux R et al: The diagnosis of dementia due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers Dement 2011,7(3):263-269; Albert MS, DeKosky ST, Dickson D, Dubois B, Feldman HH, Fox NC, Gamst A, Holtzman DM, Jagust WJ, Petersen RC et al: The diagnosis of mild cognitive impairment due to Alzheimer's disease: recommendations from the National Institute on The results were based on the Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers Dement 2011,7(3):270-279.VaD and DLB subjects were selected from the 1990 NINDS-AIREN International Workshop (Roman GC, Tatemichi TK, Erkinjuntti T, Cummings JL, Masdeu JC, Garcia JH, Amaducci L, Orgogozo JM, Brun A, Hofman A et al: Vascular dementia: diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology 1993, 43(2):250-260.) and the 4th report of the DLB Consortium (McKeith IG, Boeve BF, Dickson DW, Halliday G, Taylor JP, Weintraub D, Aarsland D, Galvin J, Attes J, Ballard CG et al: Diagnosis and management of dementia with Lewy bodies: Fourth consensus report of the DLB Consortium. Neurology 1993, 43(2):250-260.). 2017,89(1):88-100. All healthy controls had an MMSE score of ≥23.

[0044] All data in this study were obtained from publicly available sources.

[0045] Detection of miRNA expression levels Serum miRNA extraction and expression profiling were described in Shigemizu D, Akiyama S, Asanomi Y, Boroevich KA, Sharma A, Tsunoda T, Matsukuma K, Ichikawa M, Sudo H, Takizawa S et al: Risk prediction models for dementia constructed by supervised principal component analysis using miRNA expression data. Commun Biol 2019,2:77. Briefly, serum samples were isolated and transferred to a -80°C refrigerator for storage. Total RNA was then extracted and comprehensive miRNA expression analysis was performed using a human miRNA oligochip designed to detect 2,562 miRNA sequences. Then, normalization of miRNA expression was performed.

[0046] Screening for differentially expressed miRNAs First, the miRNA probes were sorted from high to low expression levels, and the top 1,000 miRNA probes were selected. To further identify the differentially expressed miRNAs, the Limma package in R software was used. According to the criteria of P<0.05 and fold change>1.3, 22 miRNAs were finally obtained.

[0047] Building a System Model By combining lasso-logistical regression and 5-fold cross-validation to build a model on the training set data, and performing validation on the validation set data, and determining the best model by scaling the negative log-likelihood (-LogL(β)), the smaller the scaled -LogL(β) value in the validation set, the better the fit of the model. Finally, the 11 miRNAs, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex were included in the final model as independent variables (see Table 1). Table 1 shows the parameter estimates, P-values, and contribution rates of each variable in the model building process.

[0048] [Table 1]

[0049] Table 1 shows that APOE4 allele status and female gender were the main risk factors for mild cognitive impairment and dementia, accounting for 53.4%, 11.1%, and 3.5%, respectively. Among the 11 miRNAs, only three, hsa-miR-6761-3p, hsa-miR-3173-5p, and hsa-miR-6716-3p, contributed most. Therefore, we removed the other eight with small contribution rates and reconstructed the model, and the modeling results are shown in Figure 1.

[0050] Based on the above method, the following Equation 1 was confirmed as a prediction model in this example. p=1 / [1+e -(i+a*Age+b*性別+c*APOE型+d*hsa-miR-6761-3p+f*hsa-miR-3173-5p+g*hsa-miR-6716-3p) ] (Formula 1) where p is the probability of dementia or early cognitive impairment; i is -19.23944, a is 0.2039311, d is 1.2992589, f is -0.398038, and g is 0.2794064. If the subject is female, the value of b is 0, and if the subject is male, b is -0.726898. If the subject does not express the APOE4 genotype, the value of c is 0; if the subject's APOE4 genotype is homozygous, the value of c is 2.4565261; and if the subject's APOE4 genotype is heterozygous, the value of c is 1.5060673.

[0051] The predictive effects of the model constructed using the above method for other dementia types and MCI are shown in Figure 2.

[0052] The model established using AD data in this application included only non-invasive parameters such as the subject's levels of three miRNAs (miR-6761-3p, miR-3173-5p, and miR-6716-3p), the subject's apolipoprotein E4 (APOE4) genotype, the subject's age, and the subject's gender. It performed well not only in predicting AD, but also in predicting VaD dementia, DLB dementia, and MCI, with an AUC of 0.874 (0.834, 0.914) for AD validation data, 0.835 (0.798, 0.872) for VaD dementia, 0.856 (0.824, 0.888) for DLB dementia, and 0.808 (0.765, 0.851) for MCI, respectively.

[0053] The grouping results of the AD validation set using the constructed model are shown in Figure 3, where AD represents patients who actually have AD and NC represents the negative control. From the results in Figure 3, we can obtain the following grouping criteria for models predicting dementia or mild cognitive impairment: If the calculated probability (p) of a subject having dementia or mild cognitive impairment is less than 10%, the subject has a low risk of having dementia or mild cognitive impairment. For example, as shown in Figure 3, if the calculated p value is 0 to 0.1, the actual incidence of AD is about 5%, which means that the risk of AD is at a low level. When the calculated probability (p) of the subject having dementia or mild cognitive impairment is 10% or more and less than 50%, the subject has a medium risk of having dementia or mild cognitive impairment. For example, as shown in FIG. 3, when the calculated p value is 0.1 to 0.2, the actual incidence of AD is 10% or less. When the calculated p value is 0.2 to 0.3, the actual incidence of AD is about 20%. When the calculated p value is 0.3 to 0.5, the actual incidence of AD is less than about 30 to 40%, which means that the subject has a certain risk of having AD. If the calculated probability (p) of a subject having dementia or mild cognitive impairment is 50% or more and less than 90%, the subject has a medium to high risk of having dementia or mild cognitive impairment. For example, as shown in Figure 3, if the calculated p value is 0.5 to 0.9, the actual incidence of AD is more than 50%, and in some cases even reaches 87%. This means that the risk of developing AD is already quite high. If the calculated probability (p) of a subject having dementia or mild cognitive impairment is 90% or more, the subject is at high risk of having dementia or mild cognitive impairment. For example, as shown in Figure 3, if the calculated p value is 0.9 to 1, the actual incidence of AD should be more than 95%. This means that the risk of such a subject is very high.

[0054] Figure 3 shows the grouping validation based on AD patients. In practical situations, AD sample size is the largest, and therefore the most representative. Therefore, the grouping criteria of the prediction model of the present application is basically confirmed based on the calculation results of Figure 3 above.

[0055] Although the embodiments of the present invention have been described in combination with the accompanying drawings, the present invention is not limited to the specific embodiments and application fields described above. The specific embodiments described above are not limiting, but merely illustrative and teaching. Under the teachings of this specification, those skilled in the art can also make many forms without departing from the scope of protection of the claims of the present invention, and all of them are included in the protection of the present invention.

Claims

1. A system for predicting dementia or mild cognitive impairment, comprising: a data collection module for acquiring data regarding the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's gender; A dementia or mild cognitive impairment probability calculation module for calculating the probability (p) that the subject suffers from dementia or mild cognitive impairment by calculating the data acquired by the data collection module; Including, the system.

2. The system according to claim 1, wherein the miRNA is one or more selected from hsa-miR-6761-3p, hsa-miR-3173-5p, and hsa-miR-6716-3p, and preferably the miRNA includes hsa-miR-6761-3p, hsa-miR-3173-5p, and hsa-miR-6716-3p.

3. The subject's apolipoprotein E4 genotype is a typing of apolipoprotein E4 alleles. The system of claim 1 .

4. The dementia or mild cognitive impairment probability calculation module pre-stores a formula for calculating the dementia probability (p) fitted by logistic regression based on the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex data in the existing database; The system of claim 1 .

5. The formula is the following Formula 1: p = 1 / [1+e －（ｉ＋ａ＊Ａｇｅ＋ｂ＊性別＋ｃ＊ＡＰＯＥ型＋ｄ＊ｈｓａ－ｍｉＲ－６７６１－３ｐ＋ｆ＊ｈｓａ－ｍｉＲ－３１７３－５ｐ＋ｇ＊ｈｓａ－ｍｉＲ－６７１６－３ｐ） ] (Formula 1) and In the formula, p is the probability of having dementia or mild cognitive impairment, APOE type is the apolipoprotein E4 genotype of the subject, and i, a, b, c, d, f, and g are unitless parameters; In the dementia probability calculation module, the values ​​of a, b, c, d, f, and g are obtained based on the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex, and are substituted into Formula 1 for calculation; The system of claim 4.

6. i is any value selected from −22.77226 to −15.70663, and is preferably −19.23944; a is any value selected from 0.1653123 to 0.2425499, and is preferably 0.2039311; d is any value selected from 0.8932064 to 1.7053113, preferably 1.2992589; f is an arbitrary value selected from −0.722253 to −0.073824, and is preferably −0.398038; g is any value selected from 0.0263939 to 0.5324189, preferably 0.2794064; The system of claim 5.

7. If the subject is female, the value of b is 0; If the subject is male, b is any value selected from −1.144378 to −0.309418, and is preferably −0.726898; The system of claim 5.

8. If the subject does not express the APOE4 genotype, the value of c is 0; When the subject's APOE4 genotype is homozygous, the value of c is any value between 1.1429478 and 3.7701044, preferably 2.4565261; When the subject's APOE4 genotype is heterozygous, the value of c is any value between 0.9740697 and 2.038065, preferably 1.5060673; The system of claim 5.

9. A grouping module is further included, which stores default dementia or mild cognitive impairment grouping parameters in advance, and groups the calculated dementia or mild cognitive impairment probability (p) of the subject based on the grouping parameters, thereby grouping the risk of the subject suffering from dementia or mild cognitive impairment. The system of claim 1 .

10. The grouping criteria prestored in the grouping module are as follows: If the calculated probability (p) of the subject having dementia or mild cognitive impairment is less than 10%, the subject has a low risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 10% or more and less than 50%, the subject is at medium risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 50% or more and less than 90%, the subject is at medium to high risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 90% or more, the subject is at high risk of having dementia or mild cognitive impairment; The system of claim 9.

11. a data collection step of obtaining data regarding the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex; a dementia or mild cognitive impairment probability calculation step of calculating the probability (p) that the subject suffers from dementia or mild cognitive impairment by calculating the data acquired in the data collection step; A method for predicting dementia or mild cognitive impairment, comprising:

12. The method according to claim 11, wherein the miRNA is one or more selected from hsa-miR-6761-3p, hsa-miR-3173-5p, and hsa-miR-6716-3p, and preferably the miRNA includes hsa-miR-6761-3p, hsa-miR-3173-5p, and hsa-miR-6716-3p.

13. The subject's apolipoprotein E4 genotype is a typing of apolipoprotein E4 alleles. The method of claim 11.

14. In the dementia or mild cognitive impairment probability calculation step, a formula for calculating the dementia probability (p) fitted by logistic regression based on the data on the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex in the existing database is stored in advance. The method of claim 11.

15. The formula is the following Formula 1: p = 1 / [1+e －（ｉ＋ａ＊Ａｇｅ＋ｂ＊性別＋ｃ＊ＡＰＯＥ型＋ｄ＊ｈｓａ－ｍｉＲ－６７６１－３ｐ＋ｆ＊ｈｓａ－ｍｉＲ－３１７３－５ｐ＋ｇ＊ｈｓａ－ｍｉＲ－６７１６－３ｐ） ] (Formula 1) and In the formula, p is the probability of having dementia or mild cognitive impairment, APOE type is the apolipoprotein E4 genotype of the subject, and i, a, b, c, d, f, and g are unitless parameters; In the dementia incidence probability calculation step, the values ​​of a, b, c, d, f, and g are obtained based on the subject's miRNA level, the subject's apolipoprotein E4 genotype, the subject's age, and the subject's sex, and are substituted into Formula 1 for calculation. The method of claim 14.

16. i is any value selected from −22.77226 to −15.70663, and is preferably −19.23944; a is any value selected from 0.1653123 to 0.2425499, and is preferably 0.2039311; d is any value selected from 0.8932064 to 1.7053113, preferably 1.2992589; f is an arbitrary value selected from −0.722253 to −0.073824, and is preferably −0.398038; g is any value selected from 0.0263939 to 0.5324189, preferably 0.2794064; The method of claim 15.

17. If the subject is female, the value of b is 0; If the subject is male, b is any value selected from −1.144378 to −0.309418, and is preferably −0.726898; The method of claim 15.

18. If the subject does not express the APOE4 genotype, the value of c is 0; When the subject's APOE4 genotype is homozygous, the value of c is any value between 1.1429478 and 3.7701044, preferably 2.4565261; When the subject's APOE4 genotype is heterozygous, the value of c is any value between 0.9740697 and 2.038065, preferably 1.5060673; The method of claim 15.

19. A grouping step is further included in which a default dementia or mild cognitive impairment grouping parameter is stored in advance, and the calculated dementia or mild cognitive impairment probability (p) of the subject is grouped based on the grouping parameter, thereby grouping the risk of the subject suffering from dementia or mild cognitive impairment. The method of claim 11.

20. The grouping criteria previously stored in the grouping step are as follows: If the calculated probability (p) of the subject having dementia or mild cognitive impairment is less than 10%, the subject has a low risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 10% or more and less than 50%, the subject is at medium risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 50% or more and less than 90%, the subject is at medium to high risk of having dementia or mild cognitive impairment; If the calculated probability (p) of the subject having dementia or mild cognitive impairment is 90% or more, the subject is at high risk of having dementia or mild cognitive impairment; 20. The method of claim 19.

Images