Target for diagnosis of alzheimer's disease and method for detecting the same

Abstract

The present application belongs to the field of medicine and medical diagnosis, and particularly relates to a diagnostic marker for Alzheimer's disease and application thereof. Specifically, the present application provides a target for diagnosing Alzheimer's disease and a detection method thereof. More particularly, the present application relates to use of a reagent for detecting expression level of UBE2D1 gene in a sample in preparation of a composition or kit for diagnosing Alzheimer's disease or predicting risk of Alzheimer's disease, wherein the reagent for detecting expression level of UBE2D1 gene in the sample comprises: a reagent for detecting mRNA level of UBE2D1 gene in the sample, a reagent for detecting ubiquitin-conjugating enzyme E2D1 level or ubiquitin-conjugating enzyme E2D1 related regulatory protein level in the sample, a reagent for detecting CpG island methylation level of UBE2D1 gene, or any combination thereof.

Description

Technical Field

[0001] This invention relates to the field of medical diagnostics, and more specifically, to diagnostic biomarkers for Alzheimer's disease and their applications. Background Technology

[0002] Alzheimer's disease (AD) is a progressive neurodegenerative disease that has become a major challenge in global public health. With the aging population, the prevalence of AD continues to rise, currently exceeding 50 million people worldwide. The disease is characterized by progressive cognitive decline, clinically manifesting as memory loss, language impairment, decreased spatial awareness, and behavioral abnormalities. Ultimately, patients lose their ability to live independently, becoming completely dependent on others, placing a heavy economic and psychological burden on society and families. Because neurological damage is irreversible, early detection and intervention are crucial. The current gold standard for AD diagnosis is cerebrospinal fluid biomarker detection and PET imaging, but these methods are highly invasive, expensive, and dependent on equipment, making them unsuitable for large-scale early screening and universal diagnosis. Finding and accurately identifying body fluid biomarkers that reflect the early progression of AD remains a significant challenge. Existing biomarkers for early diagnosis of Alzheimer's disease (such as Aβ40, Aβ42, tau, etc.) have limitations in diagnostic accuracy and batch-to-batch stability, making it difficult to meet the clinical demand for early, reliable, and low-cost screening. Summary of the Invention

[0003] To address the limitations of existing technologies such as cerebrospinal fluid biomarker detection and PET imaging due to their high invasiveness, as well as the limitations of other biomarkers in terms of accuracy and stability, this invention provides a novel target biomarker for AD diagnosis. Through detection of fluid samples and multi-dimensional evaluation, it has significant clinical value and application prospects for early screening and risk prediction of Alzheimer's disease.

[0004] The technical problem to be solved by this invention is to provide a biomarker for the early diagnosis of Alzheimer's disease. Specifically, by detecting the mRNA expression level of the UBE2D1 gene, the CpG island methylation level of the UBE2D1 gene, and the protein level of UBE2D1 in the sample, a highly sensitive and specific judgment model is constructed through multiple dimensions such as nucleic acid and protein, which improves the accuracy of clinical diagnosis and plays an important role in the study of the disease's pathogenesis and development mechanism.

[0005] Specifically, the present invention provides the following technical solutions:

[0006] 1. Use of reagents for detecting UBE2D1 gene expression levels in samples in the preparation of compositions or kits for diagnosing Alzheimer's disease or predicting the risk of Alzheimer's disease.

[0007] 2. According to the purpose described in Project 1, wherein the reagent for detecting the expression level of the UBE2D1 gene in the sample includes: a reagent for detecting the mRNA level of the UBE2D1 gene in the sample, a reagent for detecting the level of ubiquitin-conjugating enzyme E2D1 or the level of said ubiquitin-conjugating enzyme E2D1-related regulatory protein in the sample, a reagent for detecting the CpG island methylation level of the UBE2D1 gene, or any combination thereof.

[0008] 3. According to the use described in Project 2, the reagent for detecting the methylation level of CpG islands in the UBE2D1 gene includes a reagent for detecting the methylation level of CpG islands in the promoter region of the UBE2D1 gene and a reagent for detecting the methylation level of CpG islands downstream of the transcription start site of the UBE2D1 gene.

[0009] 4. According to the use described in Project 2, wherein the reagent for detecting the mRNA level of the UBE2D1 gene in a sample includes primers and / or probes for detecting the mRNA level; the primers and / or probes are selected from the group consisting of SEQ ID NO: 1-3.

[0010] 5. According to the use described in Project 3, wherein the reagent for detecting the methylation level of CpG islands in the promoter region of the UBE2D1 gene includes primers and / or probes for detecting the methylation level; the primers and / or probes are selected from the group consisting of SEQ ID NO: 7-8.

[0011] 6. According to the use described in Project 3, wherein the reagent for detecting the methylation level of the CpG island downstream of the transcription start site of the UBE2D1 gene includes primers and / or probes for detecting the methylation level; the primers and / or probes are selected from the group consisting of SEQ ID NO: 9-14.

[0012] 7. According to the use described in Project 2, wherein the ubiquitin-binding enzyme E2D1-related regulatory proteins include p53 protein, HIF-1α, OTUB1, FOXF1, YY1, Akt, or ERK protein.

[0013] 8. The use according to any one of items 1-7, wherein the expression level of the UBE2D1 gene is positively correlated with the risk of developing Alzheimer's disease, and the higher the expression level of the UBE2D1 gene, the greater the risk of developing Alzheimer's disease.

[0014] 9. The use according to any one of items 1-8, wherein the sample is a liquid sample of the subject, including whole blood, serum or plasma.

[0015] 10. The use according to any one of items 1-9, wherein,

[0016] If the mRNA level of the UBE2D1 gene in the sample is F>2, it is assessed as a high risk of Alzheimer's disease.

[0017] If the concentration of ubiquitin-conjugating enzyme E2D1 protein in a sample is ≥10 times that in a healthy control, it is considered a high risk for Alzheimer's disease; or

[0018] If the methylation level M of the CpG island in the UBE2D1 gene in the sample is <0.3, it is assessed as a high risk of Alzheimer's disease.

[0019] 11. The use according to any one of items 1-10, wherein the composition or kit is used to predict the risk of Alzheimer's disease by a multi-indicator model, wherein the input features of the multi-indicator model include the mRNA level F of the UBE2D1 gene, the level P of the ubiquitin-conjugating enzyme E2D1, and the methylation level M of CpG islands.

[0020] 12. A system for diagnosing Alzheimer's disease or predicting the risk of Alzheimer's disease, characterized in that the system includes a subject information acquisition module, and the system further includes a diagnostic module and / or a risk assessment module, wherein:

[0021] The subject information acquisition module is used to acquire detection information on the expression level of the UBE2D1 gene;

[0022] The early diagnosis and / or prognostic assessment module is used to diagnose whether the subject is an Alzheimer's disease patient based on the detection information of the expression level of the UBE2D1 gene, or to predict the risk of the subject having Alzheimer's disease based on the detection of the expression level of the UBE2D1 gene.

[0023] 13. A computer system comprising a processor and a storage device storing computer-executable code, wherein when the computer-executable code is executed at the processor, it is configured to: acquire the expression level of the UBE2D1 gene and determine, based on a calculation of the expression level of the UBE2D1 gene, whether the subject is an Alzheimer's patient or at risk of having Alzheimer's disease.

[0024] 14. The computer system according to Item 13, wherein the expression level of the UBE2D1 gene includes: the mRNA level of the UBE2D1 gene, the level of ubiquitin-conjugating enzyme E2D1 or the level of ubiquitin-conjugating enzyme E2D1-related regulatory proteins, the CpG island methylation level of the UBE2D1 gene, or any combination thereof.

[0025] 15. The computer system according to Item 14, wherein,

[0026] If the mRNA level of the UBE2D1 gene is F > 2, it is considered to be at high risk of Alzheimer's disease.

[0027] If the concentration of ubiquitin-conjugating enzyme E2D1 protein is ≥10 times higher than the concentration of ubiquitin-conjugating enzyme E2D1 protein in healthy controls, then the individual is considered to be at high risk for Alzheimer's disease; or

[0028] If the methylation level of the CpG islands in the UBE2D1 gene is less than 0.3, the individual is considered to be at high risk for Alzheimer's disease.

[0029] 16. A computer-readable medium having instructions stored thereon that, when executed by a processor, cause the processor to execute computer-executable code as defined in item 13.

[0030] Beneficial effects:

[0031] This invention provides biomarkers for the early diagnosis of Alzheimer's disease. By detecting the mRNA expression level of the UBE2D1 gene, the methylation level of CpG islands in the UBE2D1 gene promoter region, the methylation level of CpG islands downstream of the transcription start site of the UBE2D1 gene, and the level of ubiquitin-binding enzyme E2D1 or its related regulatory proteins in a sample, the risk of the sample having Alzheimer's disease can be obtained, and it can be used for screening or diagnosis of Alzheimer's disease.

[0032] This invention also constructs multiple data models based on the screened biomarkers, which can be used to screen or diagnose Alzheimer's disease. After adjusting the constants of the data models, a data model of the risk probability of developing Alzheimer's disease can be obtained. When using the data models described in this invention for screening or diagnosis, the sensitivity, accuracy, convenience and reliability are high. Attached Figure Description

[0033] Figure 1 This is a graph showing the PCR results of UBE2D1 gene expression amplification in the sample to be tested. The horizontal axis represents the PCR cycle number, and the vertical axis represents the fluorescence intensity.

[0034] Figure 2 The results of the methylation reaction are shown. The horizontal axis represents the PCR cycle number, and the vertical axis represents the fluorescence intensity. The red curve is the fully methylated control, which shows an amplification curve (peak). The green curve is the fully demethylated control, which shows no amplification signal (no peak).

[0035] Figure 3The results of the unmethylation reaction are shown. The horizontal axis represents the PCR cycle number, and the vertical axis represents the fluorescence intensity. The red curve is the fully unmethylated control, which shows an amplification curve (peak), while the green curve is the fully methylated control, which shows no amplification signal (no peak).

[0036] Figure 4 The chart displays the ROC curves on the validation set, with the horizontal axis representing specificity (false positive rate) and the vertical axis representing sensitivity (true positive rate). Specifically, the blue curve represents the ROC curve of the integrated model (i.e., a multi-omics model incorporating the target's mRNA expression level, protein content, and CpG island methylation level), with an AUC of 0.92; the red curve represents the ROC curve of the protein model (i.e., a univariate model using only the target's protein content), with an AUC of 0.89; the green curve represents the ROC curve of the mRNA model (i.e., a univariate model using only the target's mRNA expression level), with an AUC of 0.77; and the yellow curve represents the ROC curve of the methylation model (i.e., a univariate model using only the target's CpG island methylation level), with an AUC of 0.72. Detailed Implementation

[0037] Definitions and Terms

[0038] In this study, the UBE2D1 gene is located on human chromosome 10 (10q21.1) (HGNC:12474), and the protein it encodes consists of 147 amino acids. The core function of UBE2D1 is to participate in the ubiquitin-proteasome system, which is involved in protein degradation pathways, signal transduction pathways, and iron metabolism regulation.

[0039] The protein encoded by the UBE2D1 gene is ubiquitin-conjugating enzyme E2D1, or simply enzyme E2D1. It belongs to the ubiquitin-conjugating enzyme E2 family, subtype D1, and protein number H5a, also known as UbcH5a or ubiquitin-conjugating enzyme H5a.

[0040] In this document, "expression level" refers to the amount of transcriptional and / or translational products of a specific gene. In this invention, the expression level of the UBE2D1 gene can be indirectly reflected by detecting the abundance of its mRNA, the abundance of its encoded protein (ubiquitin-conjugating enzyme E2D1 or its regulatory proteins such as p53), or by detecting epigenetic markers affecting gene transcription (such as the methylation status of CpG islands). The "expression level" is relative to a predetermined reference value or the average level from a healthy control group.

[0041] In this article, the mRNA level of the UBE2D1 gene refers to the relative abundance of the transcript encoding ubiquitin-conjugating enzyme E2D1 in messenger ribonucleic acid (mRNA) extracted from a subject sample (preferably whole blood, serum, or plasma) compared to a healthy subject sample. This abundance can be measured using conventional relative quantification methods, such as 2... -ΔΔCt The value is obtained by means of the method and is denoted as F-value or E-value in this paper.

[0042] In this paper, the level of ubiquitin-conjugating enzyme E2D1 refers to the concentration (pg / ml) of ubiquitin-conjugating enzyme E2D1 (also known as UBE2D1 protein), a protein product encoded by the UBE2D1 gene, detected in subject samples (preferably serum or plasma), denoted as P-value in this paper.

[0043] In this paper, the CpG island methylation level refers to the arithmetic mean of the methylation rate of CpG islands in the UBE2D1 gene promoter region and the methylation ratio of CpG islands downstream of the UBE2D1 gene transcription start site, which is denoted as the M value in the model of this paper.

[0044] The methylation rate of CpG islands in the UBE2D1 gene promoter region is calculated as follows: Methylation rate (%) = C peak height / (C peak height + T peak height) * 100%; In practice, the arithmetic mean of the methylation rate (%) of each CpG site in the region can be taken as the methylation rate of the region. The methylation ratio of CpG islands downstream of the UBE2D1 gene transcription start site refers to the proportion of methylated alleles in the CpG islands downstream of the UBE2D1 gene transcription start site to the total number of alleles in the region, which is obtained by the following calculation: PMR (%) = [2 -ΔCt_M / (2 -ΔCt_M +2 -ΔCt_U )] × 100%.

[0045] In this paper, the "multi-index model" includes logistic regression, random forest, or support vector machine models, which use the mRNA level F of the UBE2D1 gene, the level P of the ubiquitin-conjugating enzyme E2D1, and the methylation level M of CpG islands as input features, and output the probability of having Alzheimer's disease. When using a logistic regression model, the cutoff value can be determined based on the ROC curve of the validation set using conventional methods such as the Youden index.

[0046] In this document, "sample" refers to biological material obtained from a subject. In this invention, the sample is preferably a liquid sample, such as whole blood, serum, or plasma. These samples can be obtained using non-invasive or low-invasive methods well-known in the art, such as intravenous puncture, and are readily applicable in clinical practice.

[0047] In this paper, "CpG island methylation" refers to methylation modification of cytosine residues in CpG dinucleotide-rich regions (i.e., CpG islands) of a DNA sequence. CpG island methylation in promoter regions or downstream of transcription start sites is typically associated with gene transcriptional repression. This invention unexpectedly revealed that low methylation status (i.e., low methylation rate) of CpG islands in the UBE2D1 gene promoter region and downstream of transcription start sites is significantly associated with high expression of the UBE2D1 gene and a high risk of Alzheimer's disease.

[0048] In this article, "reference value" refers to a threshold or range representing parameters such as UBE2D1 gene expression level and methylation rate in healthy individuals or individuals without Alzheimer's disease risk. This reference value can be determined by examining samples from one or more healthy control groups and using appropriate statistical methods (such as mean, median, percentile).

[0049] In this document, "risk assessment / judgment" refers to classifying subjects into different risk levels (e.g., low risk, medium risk, high risk) based on the comparison results of detected UBE2D1 gene-related parameters with preset thresholds. The specific thresholds provided in this invention (e.g., F > 2, M < 0.3) are exemplary. In practical applications, these thresholds can be corrected and optimized according to factors such as the testing platform, reagent batch, and population background, and all should fall within the equivalent scope defined by the claims of this invention.

[0050] Technical solution of the present invention

[0051] In one aspect of the invention, the use of reagents for detecting the expression level of the UBE2D1 gene in a sample in the preparation of compositions or kits for diagnosing Alzheimer's disease or predicting the risk of Alzheimer's disease is provided.

[0052] Specifically, the reagents for detecting the expression level of the UBE2D1 gene in the sample include: reagents for detecting the mRNA level of the UBE2D1 gene in the sample, reagents for detecting the level of ubiquitin-binding enzyme E2D1 or its related regulatory proteins in the sample, reagents for detecting CpG island methylation in the promoter region of the UBE2D1 gene, reagents for detecting CpG island methylation downstream of the transcription start site of the UBE2D1 gene, or any combination thereof.

[0053] Methods and reagents for detecting the mRNA level of the UBE2D1 gene in samples

[0054] The reagents used to detect the mRNA level of the UBE2D1 gene in the sample refer to a combination of reagents used to detect the expression level of the UBE2D1 gene at the mRNA level. The reagents are preferably used for detection based on real-time quantitative PCR (qRT-PCR) technology and include the following components:

[0055] 1. Sample RNA extraction reagent

[0056] The sample RNA extraction reagent described in this embodiment is used to extract high-purity, intact total RNA from samples. Depending on different experimental conditions and sample characteristics, conventional RNA extraction methods can be used, as long as they can obtain RNA samples that meet the requirements for downstream qRT-PCR detection. For example, RNA extraction can be performed using any of the following methods: lysis method, phase separation method, or magnetic bead method.

[0057] 2. After obtaining a qualified RNA sample, conventional mRNA quantification methods can be used for detection, including quantitative real-time PCR (qRT-PCR). The detection methods and corresponding reagents used in the preferred embodiment of this invention are as follows:

[0058] Based on the mRNA sequence of the human UBE2D1 gene (GenBank accession number: NM_003338.5) and the mRNA sequence of the internal reference gene GAPDH (GenBank accession number: NM_001289746.2), specific primers and TaqMan probes were designed.

[0059] As a preferred embodiment, the sequences of the primers and probes are as follows:

[0060] UBE2D1 gene:

[0061] Upstream primer (SEQ ID NO: 1): 5'-TTCCACTGGCAAGCCACTAT-3'

[0062] Downstream primer (SEQ ID NO: 2): 5'-TGTGAAAGCAATCTTTGGTGGT-3'

[0063] TaqMan Probe (SEQ ID NO: 3): 5' (FAM) - AGGACCTGAAGGCCAAGGTCA - (BHQ1) 3'

[0064] Internal reference genes are used to standardize and correct for differences between samples, thereby obtaining the true relative expression levels of the template gene. The internal reference gene used in this paper is GAPDH.

[0065] Upstream primer (SEQ ID NO: 4): 5'-CACTAGGCGCTCACTGTTCT-3'

[0066] Downstream primer (SEQ ID NO: 5): 5'-GCGCCCAATACGACCAAATC-3'

[0067] TaqMan probe (SEQ ID NO: 6): 5' (HEX)-CCGCTGGACGTCAACTGCCA-(BHQ1) 3'

[0068] It should be noted that the primer and probe sequences described above are merely illustrative examples. Those skilled in the art can design other primers and probes that can also specifically amplify and detect based on the known sequences of UBE2D1 and GAPDH, and these variants are all within the scope of protection of this invention.

[0069] In addition, the internal reference gene GAPDH is only an example for illustration, and other internal reference genes known in the art, such as ACTB, β-actin, RPL13A, etc., can also be used.

[0070] 3. Result Calculation and Result Judgment

[0071] The relative expression level of the UBE2D1 gene was calculated using conventional relative quantification methods such as the ΔCt method, Pfaffl method, and 2^(-ΔΔCt) method, with the 2^(-ΔΔCt) method being preferred.

[0072] The relative expression level of the sample is F = 2^(-ΔΔCt) (which represents the fold change in the expression of the UBE2D1 gene relative to the control sample).

[0073] F > 2, meaning that the mRNA expression level increases more than twice, indicates a higher risk of AD.

[0074] Methods and reagents for detecting CpG island methylation in the promoter region of the UBE2D1 gene.

[0075] In mammalian genomes, the methylation level of CpG islands in promoter regions is negatively correlated with gene expression; that is, the higher the methylation level, the more repressed the gene transcription. The UBE2D1 gene promoter region contains CpG islands, and their methylation status may be involved in regulating UBE2D1 gene expression. The reagents are preferably used for detection based on techniques such as methylation-specific PCR after bisulfite conversion (MSP), real-time quantitative methylation-specific PCR (qMSP), bisulfite sequencing PCR (BSP), or methylation-sensitive high-resolution melting curve analysis (MS-HRM), including the following steps and their components:

[0076] 1. Genomic DNA extraction and reagents

[0077] Used to extract genomic DNA from samples. The concentration and purity of the extracted DNA were determined by ultraviolet spectrophotometry (UV spectrophotometry). 260 / A 280 The ratio was between 1.8 and 2.0, and the integrity of the DNA was detected by 1% agarose gel electrophoresis.

[0078] 2. Bisulfite conversion and its reagents: Used to convert unmethylated cytosine (C) in genomic DNA to uracil (U), while methylated cytosine (5-mC) remains unchanged, thereby distinguishing between methylated and unmethylated CpG sites. The conversion method preferably uses a commercially available bisulfite conversion kit, which mainly includes a bisulfite solution.

[0079] 3. PCR detection and reagents: Existing conventional detection methods can be used, such as methylation-specific PCR (MSP) detection, real-time quantitative methylation-specific PCR (qMSP) detection, bisulfite sequencing PCR (BSP) detection, and methylation-sensitive high-resolution melting curve analysis (MS-HRM) detection.

[0080] In this paper, bisulfite sequencing PCR (BSP) is preferred for detection. It involves PCR amplification of bisulfite-converted DNA followed by cloning and sequencing to obtain the methylation status information of each CpG site within the CpG island region.

[0081] The BSP detection reagent contains BSP primer pairs for the UBE2D1 gene (i.e., UBE2D1-BSP-F and UBE2D1-BSP-R). These primers are designed based on the DNA sequence of the UBE2D1 gene promoter region after bisulfite conversion. The primer sequences should avoid CpG sites to ensure equal amplification efficiency for methylated and unmethylated alleles. Specifically, the primers should be designed in regions without CpG islands or use degenerate bases to cover potential CpG sites.

[0082] UBE2D1-BSP-F:TGGTTTGGTTGTGTTTTTATTTAAAT (SEQ ID NO: 7)

[0083] UBE2D1-BSP-R:AAATTATCTCACTCACCCTACACAC (SEQ ID NO: 8)

[0084] The BSP procedure is as follows: PCR amplification of the bisulfite-converted DNA is performed using BSP primers. The PCR product is identified by agarose gel electrophoresis and then purified; sequencing is then performed. The sequencing results are compared with the original UBE2D1 gene sequence, the C / T ratio of each CpG site in each clone is calculated, and the methylation frequency of CpG islands in the UBE2D1 gene is calculated.

[0085] Reagent for detecting CpG island methylation downstream of the transcription start site in the UBE2D1 gene.

[0086] Unlike traditional promoter region CpG island methylation, which primarily plays a role in transcriptional repression, genebody CpG islands (CpG islands located downstream of the transcription start site) also participate in gene expression regulation, exhibiting a different regulatory pattern than promoter region CpG island methylation. The methylation status of CpG islands downstream of the UBE2D1 gene transcription start site regulates the expression level of the UBE2D1 gene. Detection methods include methylation-specific PCR (MSP) based on bisulfite conversion, quantitative real-time methylation-specific PCR (qMSP), bisulfite sequencing PCR (BSP), or methylation-sensitive high-resolution melting curve analysis (MS-HRM). This embodiment provides a reagent combination and detection method for detecting the methylation level of CpG islands downstream of the UBE2D1 gene transcription start site, based on bisulfite-converted methylation-specific PCR (MSP) and quantitative real-time methylation-specific PCR (qMSP) technologies.

[0087] 1. Genomic DNA extraction reagent

[0088] Used for extracting genomic DNA from samples. The concentration and purity of the extracted DNA were determined by ultraviolet spectrophotometry, and the integrity of the DNA was detected by 1% agarose gel electrophoresis. DNA samples were stored at -20°C or -80°C for later use.

[0089] 2. Bisulfite Conversion and its Reagents

[0090] Bisulfite treatment is a key step in methylation detection. The principle is to convert unmethylated cytosine (C) in genomic DNA to uracil (U), while methylated cytosine (5-mC) remains unchanged. After PCR amplification, uracil is replaced by thymine (T), thus creating a sequence difference between methylated and unmethylated sequences. The preferred method is to use a commercially available bisulfite conversion kit, which mainly consists of a bisulfite solution.

[0091] 3. PCR detection and reagents: Existing conventional detection methods can be used, such as methylation-specific PCR (MSP) detection, real-time quantitative methylation-specific PCR (qMSP) detection, bisulfite sequencing PCR (BSP) detection, and methylation-sensitive high-resolution melting curve analysis (MS-HRM) detection.

[0092] In this study, methylation-specific PCR (MSP) is preferred. MSP is one of the most classic methods for detecting the methylation status of CpG islands. Specific primers are designed for methylated alleles (M primers) and unmethylated alleles (U primers) to amplify the bisulfite-converted DNA template by PCR. The methylation status is determined by the presence or absence of the amplification product. In this embodiment, MSP primers are designed for the CpG island region downstream of the transcription start site of the UBE2D1 gene.

[0093] Two pairs of specific primers were designed based on the DNA sequence of the CpG island region downstream of the UBE2D1 gene transcription start site after bisulfite conversion:

[0094] (1) Methylation-specific primers (M primers, including pre-methylation primers MF and post-methylation primers MR): These primers specifically recognize methylated CpG sites that remain C after transformation. The primer sequence should contain at least 1-2 CpG sites, and these sites should be located as close as possible to the 3' end of the primer to improve specific recognition of methylated alleles. Primer length is generally 20-28 bp, with a Tm value of 55-65℃, and the Tm value difference between the upstream and downstream primers should not exceed 2℃. The amplified fragment length is 100-300 bp, including:

[0095] Premethylation primer MF: TGTAGAGGTTAATTTTAGTTTTCGG (SEQ ID NO: 9)

[0096] Methylated primer MR: ATATAATTCCAACCCTATCCTCGA (SEQ ID NO: 10)

[0097] (2) Unmethylated specific primers (U primers, including pre-methylated primers UF and post-methylated primers UR): These primers specifically recognize unmethylated alleles where C is changed to T after transformation. The position corresponding to the CpG site in the primer is replaced with T instead of C. Other design parameters of the U primers (length, Tm value, amplified fragment length) should be matched with those of the M primers as closely as possible. These include:

[0098] Unmethylated preprime UF: TGTAGAGGTTAATTTTAGTTTTTGG (SEQ ID NO: 12)

[0099] Unmethylated primer UR: ATATAATTCCAACCCTATCCTCAAC (SEQ ID NO: 13)

[0100] Methods and reagents for detecting the level of ubiquitin-conjugating enzyme E2D1 or its related regulatory proteins in samples.

[0101] The reagents used to detect the level of ubiquitin-binding enzyme E2D1 in the sample include a combination of reagents for detecting the expression level of the enzyme encoded by UBE2D1 (i.e., UBE2D1 protein) at the protein level. These reagents can be used for detection based on techniques such as enzyme-linked immunosorbent assay (ELISA), immunohistochemistry (IHC), or Western blotting. As a preferred high-throughput and high-sensitivity detection method, this embodiment employs flow cytometry based on fluorescent microspheres. This technology conjugates specific antibodies to fluorescently encoded microspheres, and a flow cytometer performs dual-channel fluorescence detection on individual microspheres (encoding fluorescence is used to identify the target protein, and reporter fluorescence is used for quantification). This enables accurate quantitative detection of UBE2D1 protein in trace samples and supports multi-factor detection.

[0102] Its reagent composition includes:

[0103] UBE2D1 Capture Antibody-Conjugated Microspheres: Mouse anti-human UBE2D1 monoclonal antibody (affinity-purified, concentration 1 mg / mL) is covalently coupled to the surface of fluorescently encoded microspheres (such as Luminex MagPlex microspheres) via carbodiimide (EDC) or glutaraldehyde method.

[0104] Biotinylated UBE2D1 detection antibody: Rabbit anti-human UBE2D1 polyclonal antibody (affinity purified) was labeled using a biotinylation kit, with the biotin to antibody molar ratio controlled at 5-10:1. After labeling, the antibody was stored in PBS buffer containing 0.5% BSA and 0.05% sodium azide at -20°C.

[0105] Streptavidin-phycoerythrin (SA-PE): High-purity streptavidin-labeled R-phycoerythrin (4 mg / mL), stored at 4°C protected from light.

[0106] Multiplex fluorescence immunoassay based on UBE2D1 upstream and downstream regulatory network proteins

[0107] UBE2D1, an E2 enzyme, synergistically promotes the polyubiquitination of p53 with E3 enzymes, leading to its degradation by the 26S proteasome, thus negatively regulating downstream p53 protein levels. Detecting both p53 protein and mRNA levels can reflect the status of UBE2D1 to some extent. Specifically, a decrease in both p53 protein and mRNA levels corresponds to an increase in UBE2D1, thereby increasing the risk of AD. Similarly, related regulatory networks include HIF-1α and OTUB1 (downstream negative); FOXF1 and YY1 (upstream positive); and Akt and ERK (downstream positive).

[0108] The difference between multicolor fluorescence detection methods and monochromatic fluorescence lies in the use of microspheres coupled with different targets and different fluorescence to simultaneously detect different targets in the same sample. However, attention must be paid to the selection of the concentration and fluorescent groups, as well as subsequent fluorescence compensation.

[0109] Examples, but not limited to:

[0110] target Abundance Fluorescent microsphere encoding Laser / Detection Filter Akt Extremely high BUV395 (weak) 405 nm / 450±20 nm ERK Extremely high Pacific Blue (weak) 405 nm / 525±25 nm YY1 Medium and high FITC (China) 488 nm / 530±15 nm OTUB1 middle PE (bright) 488 nm / 575±15 nm HIF-1α Low PE-Cy5.5 (Extreme Brightness) 488 nm / 695±20 nm FOXF1 Extremely low APC (Extremely Bright) 633 nm / 660±10 nm

[0111] Reagent test kit

[0112] In this paper, the above reagent components can be packaged separately and assembled into an Alzheimer's disease diagnostic kit. Primers and controls should be stored at -20°C. The kit includes detailed instructions for use, covering sample processing, DNA extraction, operating procedures, and result interpretation criteria.

[0113] Computer systems and readable media

[0114] In some embodiments, the present invention includes a computer system, computer-executable code, and a computer-readable medium for implementing the biomarker data analysis and disease determination of the present invention.

[0115] In some implementation schemes, the computer system can be configured as a standalone workstation, a network server, a cloud computing platform, etc.

[0116] In some embodiments, the present invention provides a computer device including a processor and a memory having instructions stored thereon that, when executed by the processor, cause the processor to perform the steps of the method described herein, including, for example, obtaining a detection result and performing a determination step.

[0117] In some embodiments, the present invention provides a computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the steps of the method described herein, including, for example, obtaining a detection result and performing a determination step.

[0118] In some embodiments, one or more steps of the algorithm described herein may be performed using a computer program. In some embodiments, the invention includes steps executed by a computer program. In some embodiments, the invention includes a computer-readable storage medium having executable instructions stored thereon that, when executed by one or more processors, cause the one or more processors to perform one or more steps of the method of the invention.

[0119] In some implementation schemes, the functional modules of the computer-executable code may include one or more of the following modules: a data import module: importing raw Ct value data from the qPCR instrument export file; a data quality control module: performing quality control checks on the imported data, including inter-well Ct value difference analysis, control verification, outlier detection, etc.; a calculation and analysis module: executing core calculation algorithms, including F-value calculation, PMR value calculation, R-value calculation, etc.; a disease determination module: automatically determining disease subtyping based on determination criteria, outputting the determination results and the determination basis for each indicator; a report generation module: generating standardized test reports; and a data management module: storing, querying, statistically analyzing, and exporting test data, complying with medical data security and privacy protection requirements, etc.

[0120] In some implementations, the computer-readable medium can be a tangible computer-readable medium, including, for example, a solid-state drive, a hard disk drive, a USB flash drive, a flash memory card, an optical disk, a read-only memory, an erasable programmable read-only memory, etc.; or an intangible computer-readable medium, including, for example, a network-transmitted digital signal, a cloud storage remote storage medium, etc. In some implementations, the computer-readable medium can be configured as locally installed software, a web application, a mobile application, etc.

[0121] The present invention will be further described in detail below with examples of several specific embodiments, which do not constitute a limitation on the scope of protection of the present invention.

[0122] I. Screening of candidate genes

[0123] By analyzing body fluid samples from animal models that closely mirror human pathological processes and conducting multi-omics data analysis and target mining, several key targets, including UBE2D1, were precisely identified within complex molecular networks through differential analysis and network modeling. Data showed a significant correlation between changes in UBE2D1 abundance in animal body fluids and cognitive decline. This finding not only reveals the molecular evolution of AD from the central to the peripheral nervous system but also provides a scientific basis for early peripheral blood diagnosis. Furthermore, the results were subjected to a large-scale data retrospective analysis using the internationally recognized Alzheimer's Disease Neuroimaging Initiative (ADNI) human clinical database. Comparative analysis revealed significant differences in UBE2D1 expression between human AD patients and healthy individuals, fully demonstrating the authenticity and reliability of the targets and pathways described in this invention.

[0124] II. qRT-PCR detection of UBE2D1 mRNA expression level

[0125] 1. Sample RNA extraction:

[0126] Stabilizing agents containing RNase inhibitors were added to the collected plasma samples from the subjects. High-purity total RNA was obtained through lysis, phase separation, or magnetic bead extraction. Due to the small sample volume (2 ml), column extraction (i.e., solid-phase extraction based on a centrifuge column) or magnetic bead extraction was preferred for RNA extraction. NanoDrop quality control indicators: A260 / 280≈1.9-2.1, A260 / 230≥1.8. The extracted RNA was reverse transcribed into cDNA. The PrimeScript RT reagent Kit (Takara Bio, Japan) was an optional kit, and the processing was carried out according to the kit instructions.

[0127] At the same time, plasma samples were collected from healthy subjects and the same procedure was performed as a healthy control.

[0128] 2. Primer and probe design:

[0129]

[0130] (Note: If UBE2D1 and housekeeping genes need to be tested together in the same tube, the probes need to be labeled with different fluorescent markers.)

[0131] 3. Method for constructing a standard plasmid template sequence containing the target gene for quality control: clone the target region corresponding to the UBE2D1 gene into a plasmid vector (the vector can be a conventional plasmid vector such as pUC57, but there are no restrictions on it).

[0132] The target region (i.e., amplification region) sequence of the UBE2D1 gene is as follows:

[0133] Ttccactggcaagccactattatggggcctcctgatagcgcatatcaaggtggagtcttctttctcactgtacattttccgacagattatccttttaaaccaccaaagattgctttcaca(SEQ ID NO: 17)

[0134] 4. Amplification detection

[0135] Prepare the qRT-PCR reaction solution as follows: Use the standard plasmid template sequence as the template for quality control reactions; use sample cDNA as the template for clinical testing.

[0136]

[0137] Amplification program settings:

[0138]

[0139] Figure 1 The image shows the PCR results of the amplification of the UBE2D1 gene expression in the sample to be tested, with a Ct value of 0.85.

[0140] 5. Data Analysis

[0141] Method for calculating the relative expression level of UBE2D1 gene mRNA: using 2 -ΔΔCt The relative expression level of the UBE2D1 gene was calculated using this method.

[0142] Due to sample size and individual differences, it is necessary to check the internal reference. A Ct value less than 35 is considered a valid result. Furthermore, the inter-group |ΔCt| < 1 for replicate groups.

[0143] a. First, standardize to the internal reference: ΔCt = Ct_sample or healthy control – Ct_internal reference

[0144] b. Compared with healthy controls, ΔCt = ΔCt_tested_sample – ΔCt_healthy_control;

[0145] c. Relative expression level of the sample F=2 -ΔΔCt ;

[0146] d. Judgment criteria: F > 2, that is, the mRNA expression level increases more than twice, which can be considered as a high risk of AD.

[0147] 6. Optimization of internal reference genes

[0148] In this embodiment, the internal reference gene is not limited to GAPDH. GAPDH is a typical housekeeping gene and is often used as a stable baseline for detection. However, existing studies have found that it is also affected by multiple factors such as diurnal or inter-meal fluctuations. That is, no corresponding input was made for baseline fluctuations caused by individual habits.

[0149] Therefore, in this embodiment, multiple internal parameters (such as GAPDH, ACTB, β-actin, RPL13A, etc.) are also set. When calculating the Ct value, the calibration internal parameters of the sample are determined according to the following formula:

[0150] Ct_ref = -log(GEOMEAN(2^-(valid internal parameter 1), 2^-(valid internal parameter 2), 2^-(valid internal parameter 3)), 2)

[0151] For each sample, ΔCt_sample = Ct_sample – Ct_ref, where the Ct value of the control group is the average, i.e., ΔCt_control = AVERAGE(all ΔCt values ​​of the control group).

[0152] The arithmetic mean is used to estimate all internal parameters on a relatively average basis, reducing the impact of fluctuations in specific internal parameters.

[0153] 7. Correction of individual information

[0154] Furthermore, to further avoid errors caused by individual information, individual information, including values ​​such as heart rate, blood pressure, and blood sugar, can be manually input or collected in multiple segments automatically by the device. Correction factors are calculated based on machine learning to assist in real-time correction of multiple detected intrinsic parameters. The corrected intrinsic parameter baseline is output as a real-time correction factor, resulting in an intrinsic parameter baseline that is updated in real time based on individual information.

[0155] III. Sequencing detection of CpG island methylation level in the upstream promoter of the UBE2D1 gene

[0156] Upstream promoter methylation is an important and relatively stable epigenetic modification that mainly occurs on CpG islands in the promoters of target genes. It affects gene expression by altering the chemical modification state of DNA. High methylation levels usually indicate gene repression, while low methylation levels usually indicate gene activation. By detecting the methylation level of UBE2D1 in the target population, a low methylation rate indicates that UBE2D1 is in an activated state, making them more susceptible to developing early-stage Alzheimer's disease (AD).

[0157] 1. Sample cfDNA extraction: Cell-free DNA was extracted from the collected plasma samples of the subjects using column chromatography or magnetic bead method. NanoDrop quality control was performed, with A260 / 280 ≈ 1.8-2.0 and A260 / 230 ≥ 1.5.

[0158] 2. Methylation Conversion: The extracted cfDNA was processed using a bisulfite conversion column kit (e.g., EZ DNA Methylation-GoldKits, Zymo Research, USA). Bisulfite, at 50–65℃ and pH 5.0–5.5, deaminates and sulfonates unmethylated cytosine (C) to uracil sulfonate, which is then desulfonated to form uracil (U). 5-methylcytosine (5mC), due to steric hindrance from the methyl group, is not converted and remains cytosine (C). After PCR amplification, U is read as T. The "C→T" sites in the original sequence represent unmethylated sites, while the sites that "remain C" represent methylated sites. During the procedure, it is important to preserve unmethylated samples.

[0159] 3. PCR detection

[0160] In the quality control reaction, plasmid standards are used as templates (the method for constructing the standard plasmid template sequence can be found in the "Method for constructing the standard plasmid template sequence containing the target gene for quality control" section of Example 2); cfDNA is used as a template in clinical testing.

[0161] (1) Primer design: Design specific primers based on the gene promoter region:

[0162]

[0163] (2) Amplification detection

[0164] Reaction system preparation (25 μL):

[0165]

[0166] Amplification program settings, note that the hot cap temperature should be 105℃:

[0167]

[0168] PCR amplification results as follows Figure 2 As shown.

[0169] 4. Product Identification and Comparison Instructions

[0170] The presence of a single bright band at 350 bp, detected by electrophoresis, indicates that the methylated PCR product was successfully obtained.

[0171] The positive control used was fully methylated DNA after bisulfite conversion, which showed a strong band.

[0172] The negative control used was fully unmethylated DNA after bisulfite conversion, and the blank control was ddH2O.

[0173] 5. PCR product purification and sequencing

[0174] Purification of PCR products is preferably performed using KAPA Pure Beads (e.g., purchased from Beijing Pukairui Biotechnology Co., Ltd.).

[0175] Specifically, transfer all 20 μL of PCR reaction product to a new 1.5 mL nuclease-free centrifuge tube. Vortex the magnetic beaded bottle to fully suspend the product.

[0176] Add 20 μL of KAPA Pure Beads (volume ratio 1:1) after thorough suspension.

[0177] Mix thoroughly by pipetting at least 10 times and incubate at room temperature for 5 minutes. Place the tube on a magnetic rack and let it stand for 5 minutes, or until the solution is clear. Carefully remove and discard the supernatant.

[0178] Keeping the tube on the magnetic rack, add 200 μL of 80% ethanol (prepared fresh for use), let stand for 0.5–1 min, and discard the supernatant. Repeat this washing step once (for a total of two washes), keeping the tube on the magnetic rack throughout the process.

[0179] Open the lid and air dry the magnetic beads for 2 minutes, until the surface of the magnetic beads is dull but not cracked.

[0180] Remove the tube from the magnetic rack, add 22 μL of 10 mM Tris-HCl (pH 8.0-8.5) or ddH2O, mix thoroughly by pipetting, incubate at room temperature for 2 min, return to the magnetic rack and let stand for 2 min. After the solution becomes clear, carefully transfer 20 μL of supernatant (i.e., purified DNA) to a new tube.

[0181] Measure the concentration (Qubit or Nanodrop), record it, and send it to the Sanger sequencer for sequencing.

[0182] 6. Data Analysis

[0183] After sequencing the products using a Sanger sequencer, the methylation status of the corresponding sites can be obtained by comparing the data with the control group. Cytosine at non-CpG sites typically shows a T peak after PCR amplification. For CpG sites: if the sequencing result is C, the site is methylated; if the sequencing result is T, the site is not methylated. If partially methylated, the site will exhibit a "bimodal" pattern, meaning both a C peak and a T peak appear simultaneously. The methylation rate (%) at this site = C peak height / (C peak height + T peak height) * 100%.

[0184] IV. Detection of CpG island methylation level downstream of the UBE2D1 gene transcription start site

[0185] CpG islands located downstream of the transcription start site are strong markers of gene silencing. By detecting the methylation level of CpG islands downstream of the UBE2D1 start site in a sample, a low methylation rate indicates that UBE2D1 is in an activated state, making it more likely to develop into early AD.

[0186] 1. Sample cfDNA extraction: The collected liquid samples were subjected to cell-free DNA extraction using column chromatography or magnetic bead method. NanoDrop quality control was performed, with A260 / 280 ≈ 1.8-2.0 and A260 / 230 ≥ 1.5.

[0187] 2. Methylation Transformation: The extracted cfDNA was processed using a bisulfite transformation column kit (e.g., EZ DNA Methylation-GoldKits, Zymo Research, USA). Bisulfite, at 50–65 °C and pH 5.0–5.5, deaminates and sulfonates unmethylated cytosine (C) to uracil sulfonate, which is then desulfonated to form uracil (U). 5-methylcytosine (5mC), due to steric hindrance from the methyl group, is not transformed and remains cytosine (C). After PCR amplification, U is read as T. The "C→T" sites in the original sequence represent unmethylated sites, while the sites that "remain C" represent methylated sites. The degree of methylation of the original cfDNA was obtained by setting up fully methylated and fully unmethylated controls.

[0188] 3. PCR detection

[0189] When performing quality control reactions, plasmid standards are used as templates; cfDNA is used as a template for clinical testing; each sample is set up with methylation reaction wells (M), non-methylation reaction wells (U), fully methylated control reaction wells (M control) and fully non-methylated reaction wells (U control).

[0190] (1) Primer and probe design:

[0191]

[0192] Note: There are no special restrictions on the fluorescent labeling of the probe; conventional selection can be made as needed.

[0193] (2) Method for constructing standard quality plasmid template sequences containing the target gene for quality control: The fragments of the fully methylated sequence template (original sequence) and the fully unmethylated sequence template (all C bases of the original sequence are changed to T bases) of the UBE2D1 gene-related amplification region are cloned into the pUC57 vector respectively;

[0194] The fully methylated insert sequence (amplified region) of the UBE2D1 gene is as follows:

[0195] TGTAGAGGTTAATTTTAGTTTTCGGTAGTTTTTATTTTAAGTTGTTTGGTTGAGGTTAGTAGTAAGTTATTTGGTAGCGTTCGGGGGTTTGTAGTCGAGGATAGGGTTGGAATTATAT (SEQ ID NO: 15)

[0196] The fully unmethylated insert sequence (amplified region) of the UBE2D1 gene is as follows:

[0197] TGTAAGGTTTATTTTTAGTTTTTTGGTAGTTTTTATTTTAAGTTGTTTGGTTGAGGTTAGTAGTAAGTTATTTGGTAGTGTTTGGGGGTTTTGTAGTTGAGGATAGGGTTGGTATTATAT (SEQ ID NO: 16)

[0198] (3) Amplification detection

[0199] Reaction system preparation (50 μL):

[0200]

[0201] Amplification program settings:

[0202]

[0203] The criteria for a positive result are a typical S-shaped amplification curve with a Ct value < 35. PCR amplification results are as follows: Figure 3 As shown.

[0204] (4) Data Analysis

[0205] Record the Ct value for each reaction well and calculate the methylation ratio (Percentage of Methylated Reference, PMR):

[0206] Methylation Ct value: ΔCt_M = Ct(M) - Ct(M control);

[0207] Unmethylated Ct value: ΔCt_U = Ct(U) - Ct(U control);

[0208] PMR (%) = [2 -ΔCt_M / (2 -ΔCt_M +2 -ΔCt_U )] × 100%.

[0209] V. Detection of UBE2D1 protein levels

[0210] The principle of flow cytometry detection based on fluorescent microspheres is as follows: microspheres coupled with antibody proteins and fluorescence are co-incubated with liquid samples, and the status of fluorescent microspheres is detected by flow cytometry, thereby reflecting the content of the corresponding proteins in the sample.

[0211] 1. Detection method for UBE2D1 protein ubiquitin-binding enzyme E2D1

[0212] Antibody magnetic bead preparation: After activation of the magnetic beads (e.g., Dynabeads MyOne purchased from Thermo Fisher) using conventional methods, anti-human UBE2D1 monoclonal antibody (UBE2D1 Monoclonal Antibody purchased from Thermo Fisher) was added. The mixture was incubated at 37°C in the dark for 2 h, then slowly mixed on a shaker for 2 h. After blocking with 1% BSA in PBS for 30 min, the mixture was washed twice and then the concentration of the magnetic beads was determined by conventional flow cytometry to 1×10^5 beads / µL.

[0213] Centrifuge the supernatant of the collected plasma samples at 1500g for 10 min at 4°C, take 50 μL of the supernatant, add protease inhibitor according to the ratio, add 2 volumes of PBS (containing 0.5% BSA-0.05% Tween-20), add 25 μL of antibody magnetic beads, mix on a shaker at 37°C for 1 h, and wash twice.

[0214] Add 100 µL of biotinylated UBE2D1 polyclonal antibody (e.g., bs-8356R-Biotin purchased from Bioss) (2 µg / mL), mix on a shaker at 37°C for 30 min, and wash twice;

[0215] Add 100 µL of streptavidin SA-phycoerythrin PE-Cy5.5 (diluted to 0.5 µg / mL with PBS (containing 0.5% BSA-0.05% Tween-20) → incubate at room temperature in the dark for 20 min, then wash twice.

[0216] Detection was performed using a dual-laser flow cytometer (such as Luminex 200, FLEXMAP 3D, or BD CBA platform). At least 50-100 events were collected for each microsphere region. The encoding laser (e.g., 635 nm) identified the microsphere region, and the reporting laser (532 nm) detected the PE fluorescence signal (expressed as average fluorescence intensity, MFI). Concentration calibration can be performed using recombinant UBE2D1 protein solutions of known concentrations; however, due to personnel and experimental differences, it is preferable to obtain the corresponding judgment through comparison with a normal population. The logarithm of the concentration of the UBE2D1 recombinant protein standard or the sample from a healthy subject is used. 10 Using the concentration as the x-axis (X) and the measured MFI value as the y-axis (Y), a standard curve is fitted using 5-parameter logistic regression (5-PL) or 4-parameter logistic regression (4-PL) to obtain the regression equation.

[0217] Sample concentration calculation: Substitute the MFI value of the sample to be tested into the standard curve equation to calculate the concentration of UBE2D1 protein in the sample (pg / mL). For serum / plasma samples, it is directly expressed as pg / mL. Based on the above operation, the concentration of UBE2D1 protein in the healthy control sample (pg / mL) is obtained, and the multiple of the UBE2D1 protein concentration in the sample relative to the UBE2D1 protein concentration in the healthy control sample is calculated.

[0218] Judgment criteria:

[0219]

[0220] VI. Build a model to achieve non-invasive and accurate identification of early-stage Alzheimer's disease (AD).

[0221] All samples used in this article were peripheral blood samples collected from patients whose clinical diagnoses had been confirmed. The patients had completed all clinical pathological tests and provided diagnostic results, and the sample collection complied with relevant standards and requirements.

[0222] 1. Data Collection:

[0223] (1) The source of the subjects and their inclusion and exclusion criteria

[0224] The samples used in this study were peripheral blood samples collected from hospital subjects, including peripheral blood samples from the experimental group and the control group that met the core clinical criteria for AD diagnosis. Individuals with other neurodegenerative diseases, malignant tumors, severe liver or kidney dysfunction, or who had taken immunosuppressants / methylation-modifying drugs within the past 3 months were excluded. Sample collection and use complied with relevant regulations. The core clinical criteria for AD diagnosis are the "Revised Criteria for Diagnosis and Staging of Alzheimer's Disease (2024)" published by the National Institute on Aging and the Alzheimer's Association (NIA-AA). Subjects in this study were clinically diagnosed with the following symptoms or pathology: progressive cognitive or behavioral decline, with clinical assessment and neuropsychological testing confirming impairment affecting at least two cognitive domains (e.g., episodic memory, executive function, language, or visuospatial abilities); and the degree of impairment being sufficient to affect independence in daily life or lead to a decline in work capacity.

[0225] The total sample size used in this embodiment was 176 cases, including 80 cases in the early AD case group and 96 healthy control group cases.

[0226] (2) Detection indicators:

[0227] UBE2D1 mRNA expression level: The sample was detected by the method described in Example 2 and the relative expression level of the sample (i.e., the relative expression level F of the sample obtained in Example 2) was used as the model input value E;

[0228] The methylation level of CpG islands in the upstream promoter region of UBE2D1 (i.e., the methylation rate obtained in Example 3) and the methylation level of CpG islands downstream of the transcription start site (i.e., the methylation ratio PMR obtained in Example 4) were detected by the methods described in Examples 3 and 4, respectively, and the arithmetic mean was taken as the input value M.

[0229] UBE2D1 protein content: The concentration (pg / mL) of UBE2D1 protein was obtained by the method described in Example 5, and the concentration (pg / mL) was used as the model input value P.

[0230] 2. Dataset partitioning:

[0231] Stratified random sampling was used to randomly divide the total 176 samples into a training set (n=117) and a validation set (n=59) in a 2:1 ratio, ensuring that there were no statistically significant differences in age, gender, years of education, etc., between the groups. The training set was used for model building and feature selection, while the validation set was used only to evaluate the ROC-AUC performance of the final model. Specifically, the training set included 58 patients in the early stages of Alzheimer's disease (AD) and 59 healthy subjects; the validation set included 29 patients in the early stages of AD and 30 healthy subjects.

[0232] 3. Model Training:

[0233] This embodiment can employ various classification models known in the art for binary classification modeling, including but not limited to Logistic Regression, Random Forest, or Support Vector Machine. The following example uses Logistic Regression to train the model using a training set and optimizes the model's hyperparameters using grid search or random search. Specific steps are detailed below.

[0234] Feature fusion and standardization: The three omics indicators E, M and P of each sample are Z-score standardized to eliminate the difference in units, and then concatenated to form a three-dimensional feature vector [E_norm, M_norm, P_norm] as the input of the model.

[0235] Model Construction and Training: Logistic Regression and Random Forest models are selected for binary classification. The model's decision function can be expressed as:

[0236]

[0237] Where P(AD|sample) is the predicted probability that a sample has AD, σ is the Sigmoid function, w1, w2, w3 are the model weights, and b is the bias term.

[0238] In model training, the goal is to minimize the cross-entropy loss function, and the L-BFGS optimization algorithm is used for training. The cross-entropy loss function can be expressed as:

[0239]

[0240] Where N is the number of samples, and y_i is the true label of the i-th sample. The model predicts the probability for the i-th sample.

[0241] 4. Model performance evaluation based on the validation set:

[0242] The model trained above is then used to make predictions on an independent validation set. Subsequently, the true label of each sample in the validation set and the disease probability predicted by the model (i.e., P(AD|sample)) are collected to evaluate the model performance. The core discrimination criterion is the receiver operating characteristic curve (ROC curve) and its area under the curve (AUC).

[0243] The results show (see Figure 4 The multi-omics data model based on the UBE2D1 target of this invention has excellent performance, with an AUC of 0.92, and can be accurately and sensitively used for the early diagnosis of AD.

[0244] In summary, the integrated model of this invention achieved an AUC of 0.92, surpassing the performance of the best single-indicator model, demonstrating a "synergistic complementary effect" between different omics dimensions. For some patients, their pathological state may only be reflected at the level of transcriptional regulation (mRNA) or epigenetics (methylation), without significant changes at the protein level; and vice versa. By integrating this complementary information, the multi-omics model can identify heterogeneous disease subtypes or early pathological changes that single-indicator models cannot capture, thereby achieving a non-linear improvement in diagnostic accuracy. On the other hand, from a biological perspective, mRNA expression is regulated by methylation and is ultimately translated into protein, but this process involves complex post-transcriptional regulation and differences in translation efficiency. For example, in this paper, the UBE2D1 protein model (AUC=0.89) significantly outperformed the mRNA model (AUC=0.77), indicating that for this target, protein level detection may be closer to the final functional phenotype, but may also miss early cases where mRNA has changed but the protein has not due to abnormal regulatory mechanisms. By incorporating mRNA information, the integrated model effectively covers this population segment, improving the sensitivity of early diagnosis. Simultaneously, integrating methylation level information provides an upstream mechanistic explanation for the observed changes, enhancing the biological reliability of the model results.

Claims

1. Use of reagents for detecting the expression level of the UBE2D1 gene in a sample in the preparation of compositions or kits for diagnosing Alzheimer's disease or predicting the risk of Alzheimer's disease.

2. The use according to claim 1, wherein, The reagents used to detect the expression level of the UBE2D1 gene in the sample include: reagents for detecting the mRNA level of the UBE2D1 gene in the sample, reagents for detecting the level of ubiquitin-conjugating enzyme E2D1 or the level of ubiquitin-conjugating enzyme E2D1-related regulatory proteins in the sample, reagents for detecting the CpG island methylation level of the UBE2D1 gene, or any combination thereof.

3. The use according to claim 2, wherein, The reagents for detecting the methylation level of CpG islands in the UBE2D1 gene include reagents for detecting the methylation level of CpG islands in the promoter region of the UBE2D1 gene and reagents for detecting the methylation level of CpG islands downstream of the transcription start site of the UBE2D1 gene.

4. The use according to claim 2, wherein, The reagent for detecting the mRNA level of the UBE2D1 gene in the sample includes primers and / or probes for detecting the mRNA level; the primers and / or probes are selected from the group consisting of SEQ ID NO: 1-3.

5. The use according to claim 3, wherein, The reagent for detecting the methylation level of CpG islands in the promoter region of the UBE2D1 gene includes primers and / or probes for detecting methylation levels; the primers and / or probes are selected from the group consisting of SEQ ID NO: 7-8.

6. The use according to claim 3, wherein, The reagent for detecting the methylation level of the CpG island downstream of the transcription start site of the UBE2D1 gene includes primers and / or probes for detecting the methylation level; the primers and / or probes are selected from the group consisting of SEQ ID NO: 9-14.

7. The use according to claim 2, wherein, The ubiquitin-binding enzyme E2D1-related regulatory proteins include p53 protein, HIF-1α, OTUB1, FOXF1, YY1, Akt, or ERK protein.

8. The use according to any one of claims 1-7, wherein, The expression level of the UBE2D1 gene is positively correlated with the risk of developing Alzheimer's disease; the higher the expression level of the UBE2D1 gene, the greater the risk of developing Alzheimer's disease.

9. The use according to any one of claims 1-8, wherein, The sample is a liquid sample from the subject, including whole blood, serum, or plasma.

10. The use according to any one of claims 1-9, wherein, If the mRNA level of the UBE2D1 gene in the sample is F>2, it is assessed as a high risk of Alzheimer's disease. If the concentration of ubiquitin-binding enzyme E2D1 protein in a sample is ≥10 times that in a healthy control, it is considered to be at high risk of Alzheimer's disease. or If the methylation level M of the CpG island in the UBE2D1 gene in the sample is <0.3, it is assessed as a high risk of Alzheimer's disease.

11. The use according to any one of claims 1-10, wherein, The composition or kit is used to predict the risk of Alzheimer's disease using a multi-indicator model, the input features of which include the mRNA level of the UBE2D1 gene (F), the level of the ubiquitin-conjugating enzyme E2D1 (P), and the methylation level of CpG islands (M).

12. A system for diagnosing Alzheimer's disease or predicting the risk of Alzheimer's disease, characterized in that, The system includes a subject information acquisition module, and further includes a diagnostic module and / or a risk assessment module, wherein: The subject information acquisition module is used to acquire detection information on the expression level of the UBE2D1 gene; The early diagnosis and / or prognostic assessment module is used to diagnose whether the subject is an Alzheimer's disease patient based on the detection information of the expression level of the UBE2D1 gene, or to predict the risk of the subject having Alzheimer's disease based on the detection of the expression level of the UBE2D1 gene.

13. A computer system comprising a processor and a storage device storing computer-executable code, wherein when the computer-executable code is executed at the processor, it is configured to: acquire the expression level of the UBE2D1 gene and determine, based on a calculation of the expression level of the UBE2D1 gene, whether the subject is an Alzheimer's patient or at risk of having Alzheimer's disease.

14. The computer system of claim 13, wherein the expression level of the UBE2D1 gene includes: The mRNA level of the UBE2D1 gene, the level of ubiquitin-conjugating enzyme E2D1 or the level of said ubiquitin-conjugating enzyme E2D1-related regulatory proteins, the methylation level of CpG islands in the UBE2D1 gene, or any combination thereof.

15. The computer system according to claim 14, wherein, If the mRNA level of the UBE2D1 gene is F > 2, it is considered to be at high risk of Alzheimer's disease. If the concentration of ubiquitin-conjugating enzyme E2D1 protein is ≥10 times higher than the concentration of ubiquitin-conjugating enzyme E2D1 protein in healthy controls, then the individual is considered to be at high risk for Alzheimer's disease; or If the methylation level of the CpG islands in the UBE2D1 gene is less than 0.3, it is considered a high risk of Alzheimer's disease.

16. A computer-readable medium having instructions stored thereon that, when executed by a processor, cause the processor to execute the computer-executable code as defined in claim 13.

Images