Biomarkers for parkinson's or early diagnosis of parkinson's
By using ENPP5, PLXDC2, PCDH7, and TMEM120B biomarkers and 5hmC high-throughput detection technology, a LASSO regression model was constructed, which solved the problem of early non-invasive and accurate diagnosis of Parkinson's disease and achieved high sensitivity and specificity in peripheral blood samples.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- PEKING UNIVERSITY THIRD HOSPITAL (THE THIRD CLINICAL MEDICAL SCHOOL OF PEKING UNIVERSITY)
- Filing Date
- 2026-03-06
- Publication Date
- 2026-06-05
Smart Images
Figure CN122146874A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of biotechnology, and more specifically to a biomarker for the diagnosis of Parkinson's disease or early diagnosis of Parkinson's disease. Background Technology
[0002] Parkinson's disease (PD) is a prevalent and complex chronic neurodegenerative disease with a continuously rising global incidence rate. In my country, the incidence rate among people aged 65 and above is 1.7%, with over 2.5 million patients, and this number continues to grow rapidly. It is projected that by 2030, the number of Parkinson's patients in my country will reach 4.94 million, accounting for approximately half of the global total, making it a major disease that seriously threatens the health of the Chinese population and increases the social medical burden.
[0003] Early diagnosis of Parkinson's disease refers to the identification and diagnosis of Parkinson's disease in the early and preclinical stages of the disease, before patients develop typical motor symptoms or when motor symptoms are mild and have not yet caused significant functional impairment. This provides a critical window period for early intervention and delaying disease progression.
[0004] Early diagnosis of Parkinson's disease currently faces many prominent challenges in clinical practice: (1) Early symptoms are atypical and highly overlap with other neurodegenerative diseases such as progressive supranuclear palsy, essential tremor, and multiple system atrophy, making it easy to misdiagnose and delay diagnosis; when typical motor symptoms such as resting tremor, muscle rigidity, and bradykinesia appear, a large number of dopaminergic neurons in the substantia nigra of the midbrain have been irreversibly lost, and the effect of drug intervention is significantly reduced.
[0005] (2) Existing diagnostic methods rely heavily on clinical symptom assessment. Auxiliary examinations such as neuroimaging are costly and time-consuming, making them difficult to use for large-scale population screening and dynamic monitoring.
[0006] (3) Most of the biomarkers with diagnostic value are derived from cerebrospinal fluid. Sampling is invasive, and patients have poor compliance, making it difficult to achieve repeated testing and long-term follow-up. At the same time, single biomarkers have overlap in different neurodegenerative diseases and lack specificity, which limits their clinical application.
[0007] Currently, apart from cerebrospinal fluid examination, there is still a lack of mature, non-invasive liquid biopsy techniques and stable and reliable new biomarkers for many neurodegenerative diseases, including Parkinson's disease. In clinical practice, there is an urgent need for new biomarkers that can be used for peripheral blood testing and can achieve early and accurate diagnosis and differential diagnosis of Parkinson's disease. Summary of the Invention
[0008] The purpose of this invention is to provide a biomarker for the diagnosis of Parkinson's disease or its early diagnosis. The biomarker provided by this invention can effectively diagnose Parkinson's disease, with a sensitivity of 0.903 and a specificity of 0.792, and an AUC value of 0.854. Early diagnosis based on this biomarker has advantages such as safety and non-invasiveness, easy sample acquisition, high accuracy, and convenient operation, providing an accurate assessment of Parkinson's disease.
[0009] Therefore, the present invention adopts the following technical solution.
[0010] A first aspect of the present invention provides a biomarker comprising a combination of ENPP5, PLXDC2, PCDH7 and TMEM120B.
[0011] Secondly, the present invention provides uses of the biomarker, including for constructing models for diagnosing Parkinson's disease and for preparing products for diagnosing Parkinson's disease.
[0012] Thirdly, the present invention provides a model for diagnosing Parkinson's disease, wherein the input variable of the model is the content of the biomarkers described in the present invention.
[0013] Furthermore, the method for determining the content of the biomarker is 5hmC high-throughput detection; in a specific embodiment, the determination method is 5hmC-Seal.
[0014] Fourthly, the present invention provides a method for constructing a model for diagnosing Parkinson's disease, comprising the following steps: (1) Samples from multiple patients were tested to obtain 5hmC sequencing data of DNA; (2) The sequencing data obtained in step (1) are subjected to first filtering, screening and second filtering in sequence to obtain biomarkers for diagnosing Parkinson's disease; The first filtering includes: removing peak information that appears only in 10 or fewer samples; The screening process includes: comparing sequencing data from different samples using DEseq2 software, retaining 5hmC peak regions with read counts greater than or equal to 25, and obtaining differentially expressed biomarkers for 5hmC upregulation and downregulation according to log2FoldChage>=0.35 and p value<0.01; The second filtering includes: for the differentially expressed biomarkers obtained through the screening, a machine learning model for diagnosing Parkinson's disease is constructed using the LASSO regression algorithm, and nine biomarkers are selected based on their contribution to the model: PRKD2, TULP2, ALKAL2, RUFY1, HDAC9, LINC01600, EXT2, ARHGAP24, and CCDC102B; (3) Input the biomarker data obtained in step (2) into the machine learning model, train the model, store the trained model, and obtain the diagnostic model of Parkinson's disease.
[0015] Furthermore, the sample is plasma; in a specific embodiment, the sample is peripheral blood.
[0016] Furthermore, the Parkinson's patients include asymptomatic Parkinson's patients and symptomatic Parkinson's patients.
[0017] Furthermore, in the second filter, the parameters used are: estimator = xgb.train(params =list(objective = "binary:logistic", max_depth = 3, eta = 0.1) , nrounds =1000) Furthermore, in step (3), the machine learning model includes: training a LASSO regression machine learning model.
[0018] LASSO regression, short for Least Absolute Shrinkage and Selection Operator, is a data mining method that adds a penalty function to commonly used multiple linear regression to continuously compress coefficients, thereby simplifying the model and avoiding collinearity and overfitting. When the coefficient is 0, it also achieves the effect of selecting variables. This invention uses LASSO regression analysis for data mining to explore factors related to Parkinson's disease and diagnose Parkinson's disease based on these factors. Through LASSO regression analysis, the weights of independent variables can be obtained, thus revealing which factors are closely related to Parkinson's disease. Furthermore, based on these weights, a diagnosis of Parkinson's disease can be made based on these factors. For the model provided in this invention, the performance of the model was evaluated using Receiver Operating Characteristic (ROC) analysis. Using pROC to calculate the area under the curve (AUC), the model's sensitivity and specificity reached 0.903 and 0.792, respectively, with an AUC value of 0.854.
[0019] Fifthly, the present invention provides a product for diagnosing Parkinson's disease, the product diagnosing whether or not a person has Parkinson's disease based on the model for diagnosing Parkinson's disease.
[0020] 5-methylcytosine (5mCs) in DNA is an important epigenetic marker that plays a crucial role in gene expression and tumorigenesis and development. This invention utilizes high-throughput sequencing technology to discover 5-hydroxymethylcytosine biomarkers for diagnosing Parkinson's disease, enabling early and effective diagnosis of whether a patient has Parkinson's.
[0021] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention provides a biomarker for diagnosing Parkinson's disease, which can make an effective diagnosis of whether someone has Parkinson's disease in advance.
[0022] (2) This invention also provides a model for diagnosing Parkinson's disease. The model has a sensitivity of 0.903 and a specificity of 0.792, with an AUC value of 0.854, exhibiting the advantages of high specificity and high sensitivity. By applying the biomarkers and / or models described in this invention, a safe, non-invasive, and highly accurate diagnosis of whether a patient has Parkinson's disease can be achieved.
[0023] (3) According to the biomarkers and models provided by the present invention, the sample for Parkinson's diagnosis is peripheral blood, which is easy to obtain; the diagnostic method relies on high-throughput sequencing, which has high detection efficiency; the diagnostic model is reliable and the diagnostic results are highly accurate. Attached Figure Description
[0024] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. In the drawings: Figure 1 Parkinson's Queue Design and Model Building Process; Figure 2 Parameters related to the construction of the Parkinson's disease diagnostic model and the diagnostic efficacy of the model in the internal validation group; Figure 3 In the external validation group, the area under the curve (AUC) of the diagnostic model provided by this invention was calculated using pROC. Figure 4 AUC values for single biomarker diagnosis in Parkinson's disease diagnostic models (training group and internal validation group); Figure 5 AUC value for a single biomarker diagnosis in the external validation set. Detailed Implementation
[0025] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
[0026] This invention provides a model for diagnosing Parkinson's disease. The relevant parameters of nine 5hmC characteristic biomarkers in the model are shown in Table 1. Table 1 In the embodiments of the present invention, the 5hmC-Seal high-throughput sequencing method is used to sequence the samples. The 5hmC-Seal high-throughput sequencing method used in the following embodiments is explained as follows: 5hmC-Seal is a high-throughput sequencing method based on 5hmC. This method uses an improved chemical glycosylation marker combined with next-generation high-throughput sequencing technology to obtain the distribution information of 5hmC on genomic DNA.
[0027] Due to the high sensitivity of chemical labeling, the input DNA can be as low as 1-10 ng. The DNA can be fragmented genomic DNA or fragmented evDNAs. According to the requirements of next-generation sequencing, the DNA fragments are padded at both ends, and then an A tail is ligated to the 3' end. Sequencing Y-shaped adapters are ligated to both ends of each DNA fragment using AT specific ligation. These adapters contain index information that distinguishes the sample and sequences such as amplification primers. Next, the 5hmC labeling step was performed. First, UDP-6-N3-Glc was added, and under specific conditions, all 5hmC on the DNA reacted to become N3-5ghmC. Then, DBCO-PEG4-Biotin was added, and all N3-5ghmC were linked to biotin. Finally, through the efficient and specific binding of biotin-magnetic beads, all DNA fragments containing 5hmC sites were screened. After PCR amplification and purification, the 5hmC-based DNA library was constructed. The size and distribution of DNA bands in each sample were analyzed using Fragment Analyzer for quality control. After precise quantification of the library by qPCR, high-throughput sequencing was performed using an Illumina NovaSeq 6000 sequencer to obtain the base sequences of all DNA fragments in the library.
[0028] Example 1 This cohort included 221 healthy individuals, 222 Parkinson's disease (PD) patients, and 27 patients with Parkinson's-like disease (NDC). Peripheral blood samples (3-4 mL) were collected from the above 470 volunteers, and evDNAs were extracted from the plasma for 5hmC-Seal high-throughput sequencing. The sequencing throughput of each sample was 5-8 Gb, and the sequencing band size was 150 bp.
[0029] like Figure 1 As shown, the 470 samples were randomly divided into a discovery cohort (346 cases) and an external validation cohort (124 cases) for biomarker screening and validation, respectively.
[0030] Peak information of plasma evDNAs from different patients in the cohort (346 cases) was compared and filtered, with each peak appearing in at least 25 samples. Using DEseq2 software, DNA sequencing data from different samples were compared to identify 5hmC peak regions with at least 25 reads. Differentially regulated biomarkers for 5hmC were then identified based on log2 FoldChage >= 0.35 and p-value < 0.01. The 4088 most significantly different biomarkers were then selected for model construction. Subsequently, the LASSO model was run 100 times to further screen biomarkers. In each training subset, differentially methylated regions (DhMRs) appearing in at least 95% of the iterations were retained. Based on this, 10-fold cross-validation was performed 100 times per round. Finally, biomarkers observed in at least 3 rounds were selected to build the final diagnostic model in the training cohort, and this model was applied to the diagnosis of samples in the internal validation cohort. The parameter α is set to a range of 0.1 to 0.9, and the value that maximizes accuracy in the internal validation queue is selected by grid search.
[0031] Subsequently, further filtering was performed on the primary model to select the nine biomarkers with the lowest model coefficients. Figure 2 a, b); at the same time, the average error of these 9 markers is the smallest ( Figure 2 c), the nine biomarkers obtained were: PRKD2, TULP2, ALKAL2, RUFY1, HDAC9, LINC01600, EXT2, ARHGAP24 and CCDC102B.
[0032] Finally, a diagnostic model was constructed using the nine selected biomarkers. A LASSO regression model was trained, and receiver operating characteristic (ROC) analysis was used to evaluate the model's performance. The diagnostic results in the discovery cohort are as follows: Figure 2 As shown in d and 2e, the AUC value is 0.923. The diagnostic results in the external validation group are as follows... Figure 3 As shown in a-3d, the sensitivity was 0.903, the specificity was 0.792, and the AUC value was 0.854; the AUC value for differential diagnosis was 0.856.
[0033] In addition, we also analyzed the internal validation group ( Figure 4 :A, B) and external validation group ( Figure 5 In the AUC values of individual biomarker diagnoses (A, B), we found that the AUC values of individual biomarker diagnoses were significantly lower than those of the combination of 9 biomarkers.
[0034] Therefore, it can be concluded that the diagnostic model constructed based on the above-mentioned nine 5hmC characteristic biomarkers provided by the present invention can effectively diagnose whether a person has Parkinson's disease.
[0035] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.
Claims
1. A biomarker for Parkinson's disease or early diagnosis of Parkinson's disease, characterized in that, The biomarkers include a combination of PRKD2, TULP2, ALKAL2, RUFY1, HDAC9, LINC01600, EXT2, ARHGAP24, and CCDC102B.
2. The use of the biomarker as described in claim 1 in the preparation of products for the early diagnosis of Parkinson's disease.
3. A product based on a model for diagnosing Parkinson's disease, characterized in that, The input variable of the model is the content of the biomarker as described in claim 1.
4. The product as described in claim 3, characterized in that, The method for determining the content of the biomarker is 5hmC high-throughput detection; preferably 5hmC-Seal.
5. The product as described in claim 3, characterized in that, The method for constructing the model includes the following steps: (1) Samples from multiple patients were tested to obtain 5hmC sequencing data of DNA; (2) The sequencing data obtained in step (1) are subjected to first filtering, screening and second filtering in sequence to obtain biomarkers for diagnosing Parkinson's disease; The first filtering includes: removing peak information that appears only in 10 or fewer samples; The screening process includes: comparing sequencing data from different samples using DEseq2 software, retaining 5hmC peak regions with read counts greater than or equal to 25, and obtaining differentially expressed biomarkers for 5hmC upregulation and downregulation according to log2FoldChage>=0.35 and p value<0.01; The second filtering includes: for the differentially expressed biomarkers obtained through the screening, using the LASSO regression machine learning algorithm to construct a primary model for diagnosing Parkinson's disease, and selecting nine biomarkers based on their contribution to the model: PRKD2, TULP2, ALKAL2, RUFY1, HDAC9, LINC01600, EXT2, ARHGAP24 and CCDC102B; (3) Input the biomarker data obtained in step (2) into the machine learning model, train the model, store the trained model, and obtain the diagnostic model of Parkinson's disease.
6. The product as described in claim 5, characterized in that, The sample is plasma; peripheral blood is preferred.
7. The product as described in claim 5, characterized in that, The patients include asymptomatic Parkinson's patients and symptomatic Parkinson's patients.
8. The product as described in claim 5, characterized in that, In the second filter, the parameters used are: estimator= xgb.train(params = list( objective = "binary:logistic", max_depth = 3, eta= 0.1) , nrounds = 1000).
9. The product as described in claim 5, characterized in that, In step (3), the machine learning model includes: LASSO regression model.