TWI837055B - Method to predict risk of offspring with alzheimer's disease - Google Patents

Abstract

The present invention relates to a method to predict risk of offspring with Alzheimer's disease. The method uses a high-throughput Micro-Western Array(MWA) system to analyze the expression of proteins in the serum of the healthy control group, Alzheimer's patients, and non-onset Alzheimer's offspring to find out the proteins with differences, and use the proteins with differences to establish a risk prediction formula, the higher the score calculated by the risk prediction formula, the higher the risk of Alzheimer's disease offspring.

Description

Methods for predicting the risk of offspring developing Alzheimer's disease

The present invention relates to a method for predicting the risk of offspring suffering from Alzheimer's disease, and in particular to a risk prediction formula established using the expression level of a specific protein, and a method for predicting the risk of offspring suffering from Alzheimer's disease by using the risk prediction formula.

Dementia is the seventh leading cause of death in the world today, and Alzheimer's Disease (AD) is the typical form of dementia. Currently, about 47 million people around the world have been diagnosed with Alzheimer's disease. Alzheimer's disease.

Alzheimer's disease is a chronic brain disease characterized by progressive intellectual decline, memory loss, and neuronal loss. These changes are caused by the extracellular deposition of amyloid β (Aβ) and the intracellular deposition of hyperphosphorylated tau protein in the prefrontal cortex and hippocampus.

The deposition of amyloid beta activates microglia to drive inflammation of brain nerves. Activating the inflammatory pathway is one of the main causes of Alzheimer's disease. Therefore, taking nonsteroidal anti-inflammatory drugs (nonsteroidal anti-inflammatory drugs) , NSAIDs) may reduce the risk of Alzheimer's disease.

Microglia are resident macrophages in the brain, while astrocytes are the most abundant cells supporting neurons in the brain. Activated microglia and astrocytes Astrocytes secrete large amounts of pro-inflammatory cytokines, chemokines, reactive oxygen species (ROS), nitric oxide (NO), etc.; when amyloid beta is present, astrocytes Cells are activated to overexpress a variety of inflammation-related factors, such as IL-1β, TNF-α, inducible nitric oxide synthase (iNOS), and nitric oxide.

In addition, NLRP3 (protein 3 containing NOD, LLR and Pyrin domains) is mainly expressed in macrophages as a sensor molecule, and together with the adaptor protein ASC and pro-caspase-1, it forms the NLRP3 inflammasome; NLRP3 is crucial to the innate immune system. Activation of the NLRP3 inflammasome leads to the release of caspase 1-dependent pro-inflammatory factors IL-1β and IL-18, and causes Gasdermin D (GSDMD)-mediated pyroptotic cell death.

The expression of NLRP3 in the brains of Alzheimer's patients is increased, which activates the NLRP3 inflammasome, leading to a decrease in the phagocytosis of amyloid beta, which in turn increases the deposition of amyloid beta. Therefore, inhibiting the NLRP3 inflammasome can reduce tau protein. Hyperphosphorylation and deposition; in addition, it is known that α1-antichymotrypsin (ACT), IL-1β, IL-2, IL-4, IL-6, IL-10, TNF-α, High plasma levels of G-CSF, TGF-β1, interferon γ (IFN-γ), and C-reactive protein (CRP) are associated with an increased risk of Alzheimer's disease; however, , it is still unclear whether other serum proteins are related to the occurrence of Alzheimer's disease or the progression of Alzheimer's disease.

On the other hand, there is currently no very early detection method for Alzheimer's disease, and the methods currently used to detect Alzheimer's disease have considerable shortcomings. For example, neuropsychological questionnaires will be affected by the educational level of the examinee, and it will take a long time to fill in the questionnaire; using positron imaging or magnetic resonance imaging, it is necessary to wait until the disease is relatively advanced to detect whether you have Alzheimer's disease, and the spinal cord liquid testing accompanying If there is severe pain, the general public is not willing to do the test; or if the tau protein and amyloid beta in the blood are tested, because their content is rare, they need to wait until the disease occurs before testing. Furthermore, the earlier Alzheimer's disease is diagnosed, the earlier it can be treated and the better the prognosis will be.

In summary, further research is needed on other types of serum proteins that may be related to the occurrence or progression of Alzheimer's disease, and to identify serum proteins that can early predict Alzheimer's disease as effective biomarkers ( biomarker), which can predict whether you are at risk of developing the disease before it develops.

In view of the above problems, the inventors of the present invention tried to find serum proteins related to the occurrence of Alzheimer's disease or its disease progression from various types of serum proteins, and analyzed the expression levels of these related serum proteins using statistical algorithms to establish a risk prediction formula. The value obtained by the risk prediction formula is used as a prediction score, especially for predicting offspring who have already developed Alzheimer's disease but have not developed the disease; the higher the prediction score, the higher the risk of developing Alzheimer's disease.

The first purpose of the present invention is to use a high-throughput Western blot microarray system (Micro-Western Array, MWA) to detect the protein expression or phosphorylation status of different antibodies in different samples to find serum proteins related to the occurrence or progression of Alzheimer's disease; this novel proteomics technology is an effective system biology tool for studying protein signaling transduction networks and protein profiles.

MWA is an improved reverse phase array platform, which is composed of GeSim Nanoplotter array instrument, GE multiphor (two-dimensional electrophoresis instrument) and Licor Odyssey infrared scanner. MWA is a high-throughput system that can simultaneously detect changes in 24 to 96 proteins in 6 to 30 samples (specimens).

The serum protein spectra of healthy people, Alzheimer's disease patients and offspring without Alzheimer's disease are different. MWA is used to analyze serum proteins to find out the serum proteins with significant differences among the above three groups of people. Among these serum proteins with significant differences, serum proteins related to the occurrence of Alzheimer's disease or its disease progression are found to serve as biomarkers for predicting the risk of Alzheimer's disease offspring.

In addition, the second purpose of the present invention is to find out serum proteins related to the occurrence of Alzheimer's disease or its disease progression through MWA, analyze the expression levels of these related serum proteins using statistical algorithms, and then establish a risk prediction formula, and use the value obtained by the risk prediction formula as the prediction score; when the prediction score is higher, it means that the risk of suffering from Alzheimer's disease is also higher. After calculation, if 17 specific proteins are used as identification predictions, the threshold for distinguishing healthy subjects from Alzheimer's patients is 0.25. Based on the samples of actual Alzheimer's patients and healthy people, the accuracy rate after MWA analysis is 100%. On the other hand, in order to speed up the detection process, simplify the detection method and reduce the detection cost, 6 specific proteins are further selected from the above 17 specific proteins according to the weight as identification predictions, and the threshold for distinguishing healthy subjects from Alzheimer's patients is -0.1. Based on the samples of actual Alzheimer's patients and healthy people, the accuracy rate after MWA analysis can reach 98.3%.

The following will use specific embodiments and the attached drawings to explain in detail the technical features of the present invention so that people with ordinary knowledge in the relevant technical field can easily understand the purpose, technical features and advantages of the present invention.

Exemplary embodiments of the invention will be more fully understood from the following detailed description and the accompanying drawings of various embodiments of the invention, which should not be construed, however, as limiting the invention to the particular embodiments, but only for explanation and understanding.

Figures 1A to 1C are schematic diagrams of the blots obtained by analyzing the Western blot microarray system in three different groups of subjects; Figures 2A to 2C are schematic diagrams of the serum of three different groups of subjects. A heat map obtained by normalizing protein expression and converting it to a logarithmic value; Figure 3 shows three different groups of subjects: Alzheimer's disease patients, adult children without Alzheimer's disease, and healthy controls. , the serum protein spectrum obtained by normalizing and converting the average expression amount of each serum protein of the three into logarithmic values; Figure 4 shows the Alzheimer's disease patients, Alzheimer's disease patients, and Alzheimer's disease patients among the three different groups of subjects. The serum protein spectrum obtained by normalizing and converting the average expression of each serum protein of adult children without Alzheimer's disease into logarithmic values; Figure 5 shows the 17 possible parameters analyzed by the Lasso algorithm. The protein combination that predicts the risk of Alzheimer's disease, using the risk prediction formula established by its protein combination, to predict the ROC curve of the first batch of subjects; Figure 6 shows the 17 possible results analyzed by the Lasso algorithm Protein combinations that predict the risk of Alzheimer's disease, using the risk prediction formula established by the protein combinations, to predict the ROC curve of the second batch of subjects; Figure 7 shows the risk prediction formula established by selecting 6 proteins with higher weights among the 17 protein combinations that can predict the risk of Alzheimer's disease analyzed by the Lasso algorithm. The ROC curve obtained from the prediction; Figure 8 is a risk prediction formula established by selecting 6 proteins with higher weights among the 17 protein combinations that can predict the risk of Alzheimer's disease analyzed by the Lasso algorithm. , the ROC curve obtained by predicting the second batch of subjects.

In order to make the technical content, purpose of the invention and the effects achieved by the present invention more complete and clear, they are described in detail below, and please refer to the disclosed drawings and figure numbers.

Example 1: Comparison of serum protein differences

Subject screening

The Department of Neurology of Kaohsiung Municipal Datong Hospital recruited patients with Alzheimer's disease (AD), adult children of AD patients without AD (ACS), and healthy controls without relatives (HC) as subjects; among them, 30 AD patients, 30 adult children of AD patients without AD, and 31 healthy controls were selected after screening. All subjects this time are the first batch of subjects.

Alzheimer's disease is diagnosed based on the National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer's Disease and Related Disorders Association (NINCDS-ADRDA) criteria and based on the Clinical Dementia Rating (CDR) score. is 0, Dementia Determination Tool 8 (instrument of Subjects whose ascertainment of dementia 8 (AD8) score was less than 2 and who were not screened to have a family history of Alzheimer's disease were included in the healthy control group (HC).

Serum protein analysis

The subjects were drawn blood, and the blood was analyzed for serum proteins using the Western blot microarray system (MWA); the Alzheimer's patients, non-Alzheimer's adult children, and healthy controls were divided into three groups, and each group added the same healthy control as a standard for comparison between different blot results; that is, each group consisted of 10 Alzheimer's patients, 10 non-Alzheimer's adult children, and 10 healthy controls, and each group had an additional healthy control as a standard. The healthy control used as the standard was the same person.

For the serum proteins to be analyzed, we have studied various related literatures and selected serum proteins that may be related to Alzheimer's disease for analysis. The types of serum proteins are as follows: Apolipoprotein: ApoD, ApoE, ApoH

Liver x receptor (LXR) signal-related proteins: LXRα, LXRβ, ABCA1, ABCG1, SREBP1, SREBP2

Hippo-YAP signaling-related proteins: YAP, TAZ, MST1, MST2, LATS1, CTGF, CYR61

Metabolism-related proteins: c_Myc, fatty acid synthase, PPARγ

Inflammation-related proteins: COX-2, NFκB_p50

Homeobox protein: nanog

The serum of each subject was used with MWA lysis buffer (240mM triacetate, 1% SDS, 5mMEDTA and 0.5% glycerol), protease inhibitors, phosphatase inhibitors, and dithiothreitol (DTT). and 1mM Na ₃ VO ₄ lysis buffer to obtain serum protein lysate; then use the GeSim Nanoplotter array instrument to spot the serum protein lysate, and spot the serum sample of each subject into 24 wells. Among them, albumin (albumin) was used as the loading control group, and different specific antibodies were added to each well, and then electrophoresis was performed by GE multiphor (two-dimensional electrophoresis instrument), and the protein was transferred to a nitrocellulose membrane. After the membrane was blocked with 100% Licor buffer for 1 hour, a diluted antibody (rabbit or mouse antibody) was added and incubated overnight at 4°C.

Next, the membrane was washed 3 times with TBST, each time for 10 minutes. After washing, the membrane was incubated with Licor fluorescent secondary antibody in a mixture of 20% Licor buffer and 80% TBS at room temperature for 1 hour. Afterwards, the membrane was washed 3 times with TBST, each time for 10 minutes. After washing, the membrane was air-dried to obtain the blotting results.

Finally, the Licor Odyssey infrared scanner was used to scan the film. For the rabbit anti-mouse antibody, the film was scanned using light of wavelengths 680nm and 780nm in sequence. The blot using the rabbit antibody showed red, while the blot using the mouse antibody showed red. Green, then use Licor Odyssey analysis software (version 3.0) to analyze these serum proteins; in addition, use the average expression level of a specific protein in the healthy control group as the standard for the protein, and set the relative protein abundance (abundance) to 1 , and then normalize the serum protein expression of each subject to a standard level, convert the values into logarithmic values with base 2 (log ₂ ), and use them to create heatmap analysis.

result

Please refer to Figures 1A to 1C, which are schematic diagrams of blots obtained by Western blot microarray analysis of three different groups of subjects. We selected 24 antibodies to detect LXR (liver X receptor) signaling pathway proteins, apolipoproteins, and inflammation-related proteins, and used Western blot microarray (MWA) to analyze and compare the proteins in the serum of 30 Alzheimer's patients, 30 adult children without Alzheimer's disease, and 31 healthy controls. The results of all serum proteins of 24 antibodies and 91 subjects are shown in Figures 1A to 1C. As can be seen from Figures 1A to 1C, there are significant differences in the color of various serum proteins among 30 Alzheimer's patients, 30 unaffected adult children of Alzheimer's patients, and 31 healthy controls.

Next, refer to Figures 2A to 2C. Figures 2A to 2C are heat maps obtained by normalizing the serum protein expression levels of three different groups of subjects and converting them into logarithmic values; from Figure 2A to 2C As can be seen in Figure 2C, various types of serum proteins detected by MWA have differences in expression levels. The red color in each figure represents an increase in the expression level of the corresponding serum protein, while the green color represents a decrease in the expression level of the corresponding serum protein.

In addition, please refer to Figure 3, which is a serum protein spectrum obtained by normalizing and converting the average expression of serum proteins in three different groups of subjects, namely, Alzheimer's disease patients, non-ADD adult children, and healthy controls. It can be seen from Figure 3 that compared with the healthy control group (HC), the serum protein spectra of Alzheimer's disease patients (AD) and non-ADD adult children (ACS) are more similar. In the serum of AD and ACS, the expression levels of ABCA1, ABCG1, LXRβ and SREBP1 are higher than those of HC, but the expression levels of ApoD, ApoE, ApoH, c_Myc, COX2, LXRα, MST1, MST2, Nanog, PPARγ, TAZ and YAP are lower than those of HC. These results suggest that these serum proteins may be associated with the progression of Alzheimer's disease.

Refer again to Figure 4. Figure 4 shows the average expression of each serum protein in three different groups of subjects: patients with Alzheimer's disease and adult children without Alzheimer's disease. , the serum protein spectrum obtained by normalizing it and converting it into a logarithmic value; considering that the adult children without Alzheimer's disease (ACS) have not yet developed into Alzheimer's disease patients (AD), the serum of the two Protein spectra may be different, so it can be seen from Figure 4 that in the serum of AD, the expression levels of ABCG1, ApoD, ApoH, COX2, LXRα and YAP are higher than those of ACS, but the expression levels of ABCA1, ApoE, c_Myc, LATS1, MST1, The expression levels of MST2, Nanog, NFκB_p50, PPARγ and SREBP2 are lower than those in ACS, which proves that the serum protein profiles of ACS and AD are indeed different.

Example 2: Establishment and evaluation of risk prediction formula for offspring with Alzheimer’s disease

From the aforementioned Example 1, the serum protein analysis of Alzheimer's disease patients, adult children without Alzheimer's disease, and healthy controls, it can be seen that there are indeed different serum proteins among the three. There are indeed differences in expression levels; therefore, the least absolute shrinkage and selection operator (Lasso) algorithm was used to analyze these serum proteins to find protein combinations that can predict the risk of Alzheimer's disease. , and use these protein combinations to establish a risk prediction formula and obtain a prediction score.

After analyzing these serum proteins by the Lasso algorithm, the protein combinations that can predict the risk of Alzheimer's disease are the following 17 serum proteins: ABCA1, ABCG1, ApoD, ApoE, ApoH, c_Myc, COX2, LATS1, LXRα, LXRβ , MST1, MST2, Nanog, NFκB_p50, PPARγ, SREBP1 and TAZ.

Next, the first risk prediction formula was established using the above 17 serum proteins, as shown in the following formula (1): Prediction score = -0.056908210752237*ABCA1+0.526412395448416*ABCG1+0.00919248573432031*ApoD+-0.0227426840915952*ApoE+-0.0334007210928206*ApoH+-0.152849110362542*c_Myc+-0.0654914828256779*COX2+0.104008716614498*LAT S1+-0.0118491370530603*LXR_alpha+0.0738440948924043*LXR_beta+0.0246146560850935*MST1+-0.540352796884419*MST2+-0.328821536622116*Nanog+-0.080653522216878*NFkB_p50+ -0.00442425403949714*PPARr+0.0162380856310409*SREBP1+0.33847940235587*TAZ Formula (1)

Among them, the serum protein multiplied by each weight parameter in formula (1) is the expression amount of serum protein. The first prediction score obtained can be used to predict the risk of Alzheimer's disease offspring. When the first prediction score is greater than or equal to 0.25 points, it means that the Alzheimer's disease offspring has a very high risk of suffering from Alzheimer's disease; and when the first prediction score is less than 0.25 points, it means that the Alzheimer's disease offspring has no risk of suffering from Alzheimer's disease. However, the closer the first prediction score is to 0.25 points, the higher the risk of Alzheimer's disease offspring suffering from Alzheimer's disease.

Please refer to Figure 5. Figure 5 shows the 17 protein combinations that can predict the risk of Alzheimer's disease analyzed by the Lasso algorithm, and the risk prediction formula established using the protein combinations to predict the first batch of subjects. The resulting ROC curve graph. As can be seen from Figure 5, the area under the curve (AUC) is 1, which means that the prediction effect of formula (1) is excellent and can accurately predict whether the offspring of Alzheimer's disease will suffer from Alzheimer's disease. Heimer's disease.

In order to confirm whether formula (1) still has excellent predictive effect for different groups of subjects, MWA was used to analyze the second group of subjects again, and formula (1) was used to predict risk; after screening, the second group of subjects included 8 Alzheimer's disease patients, 10 adult children without Alzheimer's disease, and 12 healthy controls.

Please refer to Figure 6, which is a ROC curve obtained by using the risk prediction formula established by the Lasso algorithm to analyze the 17 protein combinations that can predict the risk of Alzheimer's disease for the second batch of subjects. As can be seen from Figure 6, the area under the curve (AUC) is 0.99, which once again proves that formula (1) still has an excellent prediction effect for different subjects.

On the other hand, in order to simplify the analysis of serum proteins and achieve the purpose of using fewer serum proteins to predict the risk of Alzheimer's disease, the types of serum proteins in the above formula (1) are narrowed down to a combination of 6 serum proteins according to the weights, namely ABCG1, c_Myc, LATS1, MST2, Nanog and TAZ. The above 6 serum proteins are used to establish a second risk prediction formula, as shown in the following formula ( 2) As shown: Prediction score = 0.526412395448416*ABCG1+-0.152849110362542*c_Myc+0.104008716614498*LATS1+-0.540352796884419*MST2+-0.328821536622116*Nanog+0.33847940235587*TAZ Formula (2)

Among them, the second prediction score calculated by formula (2), when the second prediction score is greater than or equal to -0.1 points, means that the offspring with Alzheimer's disease have a very high risk of developing Alzheimer's disease; and When the second prediction score is less than -0.1 points, it means that the children with Alzheimer's disease have no risk of developing Alzheimer's disease. However, when the second prediction score is closer to -0.1 points, it means that the children with Alzheimer's disease have no risk of Alzheimer's disease. The risk of offspring developing Alzheimer's disease is higher.

Please refer to Figure 7, which is a risk prediction formula established by selecting 6 proteins with higher weights from the 17 protein combinations that can predict the risk of Alzheimer's disease analyzed by the Lasso algorithm, and the ROC curve obtained by predicting the first batch of subjects. As can be seen from Figure 7, the area under the curve (AUC) is 0.99, indicating that the prediction effect of formula (2) is extremely good. Even if the types of serum proteins used for prediction are reduced, it can still accurately predict whether the offspring of Alzheimer's disease will suffer from Alzheimer's disease.

Similarly, in order to reconfirm whether Formula (2) still has excellent prediction effect for different batches of subjects, Formula (2) is then used to predict the aforementioned second batch of subjects.

Please refer to Figure 8. Figure 8 is a risk prediction formula established by selecting 6 proteins with higher weights among the 17 protein combinations that can predict the risk of Alzheimer's disease analyzed by the Lasso algorithm. The ROC curve obtained by predicting the second batch of subjects. As can be seen from Figure 8, the area under the curve (AUC) is 1. This result can once again prove that formula (2) does have excellent predictive effects for different subjects.

In summary, the method of predicting the risk of Alzheimer's disease in offspring according to the present invention has excellent prediction effect. This method can be used to predict the risk of Alzheimer's disease in advance, based on the serum protein, before the onset of Alzheimer's disease. Different expression levels are used to predict whether the offspring with Alzheimer's disease is at risk of developing Alzheimer's disease. With conventional techniques, diagnosis can only be made after the onset of Alzheimer's disease, or even after the onset of Alzheimer's disease. Alzheimer's disease can only be diagnosed after many mutations; the earlier Alzheimer's disease is discovered and treated, the better the treatment effect will be. Therefore, the method provided by the present invention can allow subjects to treat Alzheimer's disease before it develops. Detecting the possibility of Alzheimer's disease and providing early treatment can prevent Alzheimer's disease or achieve a better prognosis.

Through the description of the above embodiments, the operation and use of the present invention and the effects of the present invention can be fully understood. However, the above embodiments are only preferred embodiments of the present invention and should not be used to limit the implementation of the present invention. Within the scope of the present invention, that is, simple equivalent changes and modifications based on the patent application scope and invention description content of the present invention are all within the scope of the present invention.

Claims (6)

A method for predicting the risk of an offspring suffering from Alzheimer's disease comprises: drawing blood from an offspring suffering from Alzheimer's disease, analyzing the blood of the offspring suffering from Alzheimer's disease by using a Western blot microarray system (MWA) to obtain the expression levels of a plurality of serum proteins; using a computer software to bring the expression levels of the plurality of serum proteins into a first risk prediction formula for calculation to obtain a first prediction score, wherein when the first prediction score is greater than or equal to 0.25 points, it means that the offspring suffering from Alzheimer's disease has an extremely high risk of suffering from Alzheimer's disease; when the first prediction score is less than 0.25 points, it means that the offspring suffering from Alzheimer's disease has no risk of suffering from Alzheimer's disease. risk; and the closer the first prediction score is to 0.25 points, the higher the risk of the offspring with Alzheimer's disease is, wherein the plurality of serum proteins are a combination of ABCA1, ABCG1, ApoD, ApoE, ApoH, c_Myc, COX2, LATS1, LXRα, LXRβ, MST1, MST2, Nanog, NFκB_p50, PPARγ, SREBP1 and TAZ; and the plurality of serum proteins are analyzed using the least absolute compression selection mechanism (Lasso) algorithm to obtain a protein combination that can predict the risk of Alzheimer's disease. The method of claim 1, wherein the plurality of serum proteins are one of ABCA1, ABCG1, ApoD, ApoE, ApoH, c_Myc, COX2, LATS1, LXRα, LXRβ, MST1, MST2, Nanog, NFκB_p50, PPARγ, SREBP1 and TAZ. When the combination is used as a protein combination that can predict the risk of Alzheimer's disease, the first risk prediction formula is as shown in formula (1), Formula (1): Predicted score =-0.056908210752237*ABCA1+0.526412395448416*ABCG1+0.00919248573432031*ApoD+-0.0227426840915952*ApoE+-0.0334007210928206*Apo H+-0.152849110362542*c_Myc+-0.0654914828256779*COX2+0.104008716614498*LATS1+-0.0118491370530603*LXR_alpha+0.0738440948924043*LXR_beta+ 0.0246146560850935 *MST1+-0.540352796884419*MST2+-0.328821536622116*Nanog+-0.080653522216878*NFkB_p50+-0.00442425403949714*PPARr+0.0162380856310409*SRE BP1+0.33847940235587*TAZ Among them, the serum protein multiplied by each weight parameter in the formula (1) is the ratio of the serum protein expression amount. The method as described in claim 2, wherein according to each weight parameter of the formula (1), a combination of ABCG1, c_Myc, LATS1, MST2, Nanog and TAZ is further selected from the plurality of serum proteins as another predictable combination. Protein combinations at risk for Alzheimer's disease. The method as described in claim 3, wherein the plurality of serum proteins is a combination of ABCG1, c_Myc, LATS1, MST2, Nanog and TAZ, which is another protein combination that can predict the risk of developing Alzheimer's disease, and is substituted into a second risk prediction formula to obtain a second prediction score, the second risk prediction formula is shown in formula (2), formula (2): prediction score = 0.526412395448416*AB CG1+-0.152849110362542*c_Myc+0.104008716614498*LATS1+-0.540352796884419*MST2+-0.328821536622116*Nanog+0.33847940235587*TAZ, where the serum protein multiplied by each weight parameter in formula (2) is the expression level of the serum protein. The method as described in claim 4, wherein when the second prediction score is greater than or equal to -0.1 points, it means that the Alzheimer's disease offspring has a very high risk of developing Alzheimer's disease; when the second prediction score When the second predicted score is less than -0.1 points, it means that the offspring with Alzheimer's disease has no risk of developing Alzheimer's disease; and when the second predicted score is closer to -0.1 points, it means that the offspring with Alzheimer's disease is at risk of developing Alzheimer's disease. The risk of Alzheimer's disease is higher. The method according to any one of claims 1 to 5, wherein the Alzheimer's disease offspring are Alzheimer's disease offspring that have not yet developed Alzheimer's disease.