CN120536571A

CN120536571A - System, device or medium for diagnosing or predicting Alzheimer's disease based on a combination of gene markers

Info

Publication number: CN120536571A
Application number: CN202511046180.1A
Authority: CN
Inventors: 方超; 张瑞静; 蒋婷婷; 黎倩; 杨沛; 蒋林
Original assignee: Hangzhou Lianchuan Biotechnology Co ltd
Current assignee: Hangzhou Lianchuan Biotechnology Co ltd
Priority date: 2025-07-22
Filing date: 2025-07-29
Publication date: 2025-08-26
Anticipated expiration: 2045-07-29
Also published as: CN120536571B

Abstract

This application discloses a system, device, or medium for diagnosing or predicting Alzheimer's disease based on a combination of gene markers, belonging to the field of tumor marker technology. The gene marker combination includes TMEM74, CBLN4, SYDE2, SLFNL1, PITPNM3, MPPED1, SYNDIG1L, ADAMTS5, BMX, GLUL, KRT79, TULP2, PRSS16, IL25, KCNJ16, SLC22A25, CRYAA, DLX2, SHISA6, CRYGA, EFCAB3, GMFB, RIN1, DMRTA1, CER1, CYP26B1, REG3A, HTRA3, SLC2A4, CRYGB, and SOX3. Using the expression level data of the gene marker combination of this application, a machine learning model is constructed that can be used to diagnose whether a subject has Alzheimer's disease or predict whether a subject is at risk of developing Alzheimer's disease, which has great clinical application value.

Description

System, device or medium for diagnosing or prognosticating Alzheimer's disease based on gene marker combinations

Technical Field

The application relates to the technical field of gene markers, in particular to a system, equipment or medium for diagnosing or predicting Alzheimer's disease based on gene marker combination.

Background

Alzheimer's Disease (AD), a progressive neurodegenerative Disease, is the most common type, accounting for about 60-70% of all cases. The disease is mainly manifested by progressive memory loss, cognitive decline and behavioral and psychological symptoms, which seriously affect the quality of life of patients.

In recent years, transcriptome sequencing technology has been widely used in the study of Alzheimer's disease as a powerful tool for studying gene expression patterns. Scientists hope to be able to reveal key molecular pathways leading to neuronal damage and death, thereby providing a theoretical basis for developing new therapeutic approaches. Along with the progress of leading-edge technologies such as transcriptomics, people's knowledge of this complex disease is continually deepening, and more effective interventions are expected to be promoted in the future.

With the rapid development of transcriptomics methods, non-invasive Alzheimer's disease detection methods based on transcriptome sequencing increasingly show great application potential, but currently there is still a lack of clinically useful Alzheimer's disease detection markers.

Disclosure of Invention

In order to solve at least one of the above technical problems, the technical scheme adopted by the application is as follows.

The first aspect of the present application provides the use of an expression level detection reagent of a gene marker combination comprising TMEM74、CBLN4、SYDE2、SLFNL1、PITPNM3、MPPED1、SYNDIG1L、ADAMTS5、BMX、GLUL、KRT79、TULP2、PRSS16、IL25、KCNJ16、SLC22A25、CRYAA、DLX2、SHISA6、CRYGA、EFCAB3、GMFB、RIN1、DMRTA1、CER1、CYP26B1、REG3A、HTRA3、SLC2A4、CRYGB and SOX3 in the preparation of a kit for diagnosing or prognosticating alzheimer's disease.

In the present application, the diagnosis is an auxiliary diagnosis, and diagnosis needs to be performed in combination with other clinical indexes. If the subject has not yet exhibited other clinical characteristics of Alzheimer's disease, the subject is deemed to be at risk for developing Alzheimer's disease and medical intervention is required.

In some embodiments of the application, the expression level refers to the mRNA level of the corresponding gene.

In some embodiments of the application, the corresponding detection reagents when the expression levels are obtained using different methods are as follows:

when the expression level is obtained by using a qRT-PCR method, the detection reagent comprises an RNA extraction reagent, a reverse transcription reagent, and amplification primers and internal reference gene primers of each gene combined by the gene markers;

when the expression level is obtained using a microarray method, the detection reagent includes an RNA labeling reagent, a hybridization reagent for combining the genes of the gene markers, and a chip scanning reagent;

When the expression level is obtained using an RNA sequencing method, the detection reagent includes a capture reagent, a library construction reagent, and a sequencing reagent for mRNA of each gene of the gene marker combination.

In other embodiments of the application, the expression level refers to the protein level of the corresponding gene.

In some embodiments of the application, the detection reagent comprises an antibody to each gene-corresponding protein of the gene marker combination.

In a second aspect, the present application provides a method of constructing a model for diagnosing or predicting Alzheimer's disease, comprising the steps of:

s1, obtaining expression level data of any one of the gene marker combinations according to the first aspect of the application in a biological sample of a population, wherein the population comprises Alzheimer 'S disease patients and non-Alzheimer' S disease subjects;

s2, constructing a machine learning model by using the expression level data obtained in the step S1.

In some embodiments of the invention, the method further comprises:

And S3, in the verification set, verifying the machine learning model obtained in the step S2.

In some embodiments of the application, the machine learning model is selected from any one of the following:

Logistic regression models, support vector machine models, decision tree models, random forest models, neural network models, XGBoost models, linear discriminant analysis models, GBDT models, ADABoost models, naive bayes models, catBoost models, lightGBM models, MLP models, and ETC models.

In a third aspect, the application provides a system for diagnosing or prognosing Alzheimer's disease comprising the following modules:

a data input module for inputting expression level data of a gene marker combination in an obtained biological sample of a subject, the gene marker combination comprising TMEM74、CBLN4、SYDE2、SLFNL1、PITPNM3、MPPED1、SYNDIG1L、ADAMTS5、BMX、GLUL、KRT79、TULP2、PRSS16、IL25、KCNJ16、SLC22A25、CRYAA、DLX2、SHISA6、CRYGA、EFCAB3、GMFB、RIN1、DMRTA1、CER1、CYP26B1、REG3A、HTRA3、SLC2A4、CRYGB and SOX3;

A storage module for storing expression level data of the gene marker combination in a biological sample of a population, the population comprising alzheimer's disease patients and non-alzheimer's disease subjects;

And the disease prediction module is respectively connected with the data input module and the database storage module and is used for constructing a machine learning model by using the expression level data of the gene marker combination in the biological samples of the population and diagnosing whether the subject has Alzheimer's disease or predicting whether the subject is at risk of having Alzheimer's disease based on the expression level data of the gene marker combination in the biological samples of the subject obtained from the data input module.

In some embodiments of the application, the detection reagent is used to detect the expression level of each gene in the gene marker combination.

A fourth aspect of the application provides a computer device comprising a memory for storing a computer program, and a processor for implementing the steps of the method according to any of the second aspects of the application when the computer program is executed.

A fifth aspect of the application provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of the method according to any of the second aspects of the application.

Compared with the prior art, the application has the following beneficial effects:

The expression level data of the gene marker combination is utilized to construct a machine learning model, so that the method can be used for diagnosing whether a subject suffers from Alzheimer's disease or predicting whether the subject suffers from Alzheimer's disease or not, and has great clinical application value.

It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the application or to delineate the scope of the application. Other features of the present application will become apparent from the description that follows.

Drawings

The above, as well as additional purposes, features, and advantages of exemplary embodiments of the present application will become readily apparent from the following detailed description when read in conjunction with the accompanying drawings. Several embodiments of the present application are illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which:

FIG. 1 shows the accuracy curves for training and cross-validation using LightGBM models in example 2 of the present application;

FIG. 2 shows cross entropy loss curves for training and cross-validation using LightGBM models in example 2 of the present application;

FIG. 3 shows the ROC curves in independent validation sets using LightGBM models in example 3 of the present application;

FIG. 4 shows a confusion matrix classified in independent validation sets using LightGBM models in embodiment 3 of the present application.

Detailed Description

Unless otherwise indicated, implied from the context, or common denominator in the art, all parts and percentages in the present application are based on weight and the test and characterization methods used are synchronized with the filing date of the present application. Where applicable, the disclosure of any patent, patent application, or publication referred to in this application is incorporated by reference in its entirety, and the equivalent patents to those patents are also incorporated by reference, particularly as they disclose definitions of pertinent terms in the art. If the definition of a particular term disclosed in the prior art is inconsistent with any definition provided in the present application, the definition of the term provided in the present application controls.

In order to make the technical problems, technical schemes and beneficial effects solved by the application more clear, the application is further described in detail below with reference to the embodiments.

The following examples are presented herein to demonstrate preferred embodiments of the present application. It will be appreciated by those skilled in the art that the techniques disclosed in the examples which follow represent techniques discovered by the inventor to function in the practice of the application, and thus can be considered to constitute preferred modes for its practice. Those of skill in the art should, in light of the present disclosure, appreciate that many changes can be made in the specific embodiments which are disclosed and still obtain a like or similar result without departing from the spirit or scope of the application.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs, the disclosure of which is incorporated herein by reference as is commonly understood by reference.

Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the claims.

The experimental methods in the following examples are conventional methods unless otherwise specified. The apparatus used in the examples described below, unless otherwise specified, were all laboratory conventional apparatus, and the test materials used in the examples described below, unless otherwise specified, were all purchased from conventional biochemical reagent stores.

Example 1 marker screening

1. Study cohort

Study cohorts were from the first medical university Xuan Wu hospital neurology, including 198 patients with 4 different levels of alzheimer's disease (48 mild, 50 moderate, 50 severe) and 50 non-alzheimer's disease subjects (all normal healthy persons).

2. Transcriptome sequencing

Transcriptome sequencing was performed using blood samples from both Alzheimer's disease patients and non-Alzheimer's disease subjects.

Sequencing libraries were prepared using VAHTS Total RNA-Seq library preparation kit (NR 603-02, vazyme). Paired sequencing was performed on an Illumina HiSeq 2000 sequencer. The sequencing process was controlled by Illumina data acquisition software.

And obtaining the gene expression level through analysis, and carrying out normalization treatment.

3. Data analysis

Screening gene markers by using a Boruta algorithm:

The Boruta algorithm is a method for feature selection based on random forests. It is mainly used to identify which features in the dataset are important and may exclude features that are not important or redundant.

The selection of features using the Boruta algorithm is very useful for improving the performance and interpretability of the machine learning model, because it can reduce the number of input features, thereby avoiding overfitting and speeding up the training process, setting n_ estimators =100 for Boruta, setting other parameters as algorithm default parameters, and screening out 31 feature genes, as shown in table 1.

TABLE 1 information on 31 characteristic genes

The signature genes in table 1 can be used as markers for diagnosing or predicting alzheimer's disease, and by detecting the expression level of the signature genes, it is obtained whether a subject suffers from or is at risk of suffering from alzheimer's disease.

EXAMPLE 2 construction of Alzheimer's disease diagnosis or prediction model

All samples were randomly divided into two groups, one group was training set (0.8,198 cases) and one group was testing set (0.2,50 cases), each group included Alzheimer's disease patient sample and non-Alzheimer's disease subject sample, and a model was constructed using the expression level data of the gene markers screened in example 1.

The inventors established a diagnosis model for Alzheimer's disease using different machine learning models, namely a logistic regression model, a support vector machine model, a decision tree model, a random forest model, a XGBoost model, a linear discriminant analysis model, a GBDT model, a ADABoost model, a naive Bayesian model, a CatBoost model, a LightGBM model, an MLP model and an ETC model, respectively, and carried out model evaluation by five-fold cross validation (5 fold cross-evaluation), and evaluated model classification efficacy using the area under the curve (Area Under the Curve, AUC) of a subject work characteristic curve (Receiver Operating Characteristic Curve, ROC), and the results are shown in Table 2.

TABLE 2 Classification efficacy of different machine learning models

As can be seen from Table 2, the LightGBM model has the best classification performance, and the decision tree model, GBDT model and CatBoost model have better classification performance.

The accuracy curves and cross entropy loss curves for training and validation using LightGBM model are shown in figures 1 and 2, respectively. As can be seen from fig. 1 and 2, as the number of training set samples increases, the verification set accuracy tends to be smooth, and cross entropy loss is minimized.

In conclusion, the LightGBM model is constructed based on the expression level of the gene marker obtained by screening, so that the patients with Alzheimer's disease can be accurately distinguished, and the method can be used for auxiliary diagnosis of Alzheimer's disease.

Example 3 independent verification of model

To verify the performance of the LightGBM model constructed in example 2, the inventors further collected independent 10 patients with alzheimer's disease and 40 subjects without alzheimer's disease, respectively obtained the expression levels of the gene markers screened in example 1, and classified using the model.

The ROC curve and classification result confusion matrix are shown in fig. 3 and 4, respectively.

The performance data of the LightGBM model constructed in the independent validation set obtained by combining fig. 3 and 4 is shown in table 3. As can be seen from FIG. 4, the LightGBM model was used for prediction, the number of True Positives (TP) was 9, the number of False Positives (FP) was 2, the number of True Negatives (TN) was 38, and the number of false negatives (FALSE NEGATIVE, FN) was 1.

Therefore, the accuracy, the precision and the recall rate of the built LightGBM model in the independent verification set are calculated.

(1) Accuracy (accuracy)

In evaluating the results of model predictions, the correct ratio is predicted.

Accuracy = (tp+tn)/(tp+fp+tn+fn) = (9+38)/(9+2+38+1) =47/50=0.94

(2) Accuracy rate (precision)

The correct ratio was predicted in samples used to comment on the model prediction as Alzheimer's disease.

Accuracy rate=tp/(tp+fp) =9/(9+2) =9/11=0.818

(3) Recall (recall)

The sample for evaluating the model prediction is Alzheimer's disease, and the sample accounts for the proportion of the actual Alzheimer's disease number of the sample.

Recall = TP/(tp+fn) = 9/(9+1) = 9/10 = 0.9

(4) F1 fraction

The F1 score is the harmonic mean of the precision and recall.

The above results are summarized and the performance in independent validation sets using the LightGBM model constructed is shown in table 3.

TABLE 3 Performance in independent validation set using built LightGBM models

In summary, example 2 still has very good discrimination performance using the LightGBM model constructed in independent validation data based on the screening of the obtained signature genes.

All documents mentioned in this disclosure are incorporated by reference in this disclosure as if each were individually incorporated by reference. Further, it will be appreciated that various changes and modifications may be made by those skilled in the art after reading the above teachings, and such equivalents are intended to fall within the scope of the application as defined in the appended claims.

Claims

1. Use of a reagent for detecting the expression level of a gene marker combination for preparing a kit for diagnosing or predicting alzheimer's disease, characterized in that the gene marker combination comprises TMEM74、CBLN4、SYDE2、SLFNL1、PITPNM3、MPPED1、SYNDIG1L、ADAMTS5、BMX、GLUL、KRT79、TULP2、PRSS16、IL25、KCNJ16、SLC22A25、CRYAA、DLX2、SHISA6、CRYGA、EFCAB3、GMFB、RIN1、DMRTA1、CER1、CYP26B1、REG3A、HTRA3、SLC2A4、CRYGB and SOX3.

2. The use according to claim 1, wherein the expression level is the mRNA level of the corresponding gene.

3. Use according to claim 2, wherein the expression levels are obtained by different methods with the corresponding detection reagents as follows:

4. The use according to claim 1, wherein the expression level refers to the protein level of the corresponding gene.

5. The use according to claim 4, wherein the detection reagent comprises antibodies to the gene markers in combination with the corresponding proteins of the respective genes.

6. A method of constructing a model for diagnosing or prognosticating alzheimer's disease, comprising the steps of:

S1, obtaining expression level data of the gene marker combination of claim 1 in a biological sample of a population, the population comprising Alzheimer 'S disease patients and non-Alzheimer' S disease subjects;

7. The method of claim 6, wherein the machine learning model is selected from any one of:

8. A system for diagnosing or prognosing alzheimer's disease comprising the following modules:

9. A computer device, comprising:

A memory for storing a computer program;

a processor for implementing the steps of the method according to any of claims 6-7 when said computer program is executed.

10. A computer-readable storage medium comprising,

The computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps of the method according to any of claims 6-7.