Correlation analysis and clinical validation of Parkinson's disease and epigenetic factor-related genes based on transcriptome data

Correlation analysis and clinical validation of Parkinson's disease and epigenetic factor-related genes based on transcriptome data | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Correlation analysis and clinical validation of Parkinson's disease and epigenetic factor-related genes based on transcriptome data Guiyue Meng, Yunqi Sun This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7376065/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Parkinson's disease (PD) is a neurodegenerative disorder, and its early diagnosis and treatment remain significant challenges. Identifying molecular biomarkers associated with PD pathogenesis is critical for disease intervention. Methods The GSE7621 and GSE49036 datasets were integrated to identify differentially expressed genes (DEGs) between PD patients' substantia nigra tissues and normal samples. Weighted Gene Co-expression Network Analysis (WGCNA) was employed to screen PD-related module genes. DEGs were cross-analyzed with epigenetic factor-related genes to identify key genes. The predictive efficacy of these key genes was evaluated using 90 machine learning models, and their expression was validated via independent GEO datasets (GSE20141, GSE42966) and reverse transcription-quantitative polymerase chain reaction (RT-qPCR). CIBERSORT was utilized to analyze immune cell infiltration, and Gene Set Enrichment Analysis (GSEA) was performed to explore the biological functions and molecular mechanisms of key genes. Results A total of 78 DEGs were identified, and 550 PD-related genes were screened through WGCNA, yielding 12 key genes. Machine learning models revealed ACTL6B, PRMT8, NAP1L2, BABAM1, and EID2 as core predictive genes for PD. Independent validation and RT-qPCR confirmed significant downregulation of ACTL6B, PRMT8, NAP1L2, BABAM1, and EID2 in PD patients. Immune infiltration analysis demonstrated altered infiltration abundance of CD8 + T cells, macrophage subtypes, and other immune cells in PD patients, with key genes linked to immune microenvironment regulation. GSEA indicated that these genes participate in pathways such as cellular metabolic reprogramming and synaptic transmission. Conclusion This study systematically identified PD-associated epigenetic regulatory genes and revealed their connections to immune microenvironment dynamics and molecular pathways, offering novel insights for early PD diagnosis and therapeutic target development. Parkinson’s disease differentially expressed genes machine learning immune infiltration epigenetic regulation Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction Parkinson's disease (PD) is a chronic and progressive neurodegenerative disorder, mainly manifested by the selective loss of dopaminergic neurons in the substantia nigra pars compacta and the abnormal aggregation of α-synuclein to form Lewy bodies [ 1 ] , resulting in core motor symptoms such as bradykinesia, resting tremor, and muscle rigidity. Typical neurodegenerative features of PD include progressive apoptosis of dopaminergic neurons, synaptic dysfunction, and a widespread neuroinflammatory response [ 2 ] . Current clinical diagnosis is mainly dependent on the appearance of motor symptoms, at which time about 50%-70% of dopaminergic neurons have been los t[ 3 ] . Existing therapies lack drugs that target core pathological mechanisms such as epigenetic inheritance or protein misfolding. The discovery of reliable biomarkers can improve patient outcomes by breaking bottlenecks in early diagnosis and accelerating the development of targeted therapies [ 4 ] . Epigenetics regulates gene expression through mechanisms such as DNA methylation, histone modification and non-coding RNA, affecting cell function and phenotype [ 5 ] , and is an important bridge connecting heredity, environment and disease. In Parkinson's disease (PD), epigenetic regulation abnormalities have been widely reported. For example, DNA methylation is significantly altered in the enhancer region of neurons in PD patients and may be associated with dysregulation of alpha-synuclein (SNCA) gene expression [ 6 ] . Abnormalities in histone modifications, such as H3K9me2 and H3K27me3, have also been found to be associated with synaptic damage and neurodegeneration [ 7 ] . Epigenetic disorders accelerate the progression of PD by mediating the interaction between environmental and genetic factors. For example, SNCA gene hypomethylation increases α-synuclein expression and promotes pathological aggregation [ 8 ] . Mitochondrial genome epigenetic variation is associated with the speed of motor and cognitive decline in PD patients [ 9 ] . These studies reveal the multi-dimensional role of epigenetics in the whole process of PD disease. In this study, multiple omics data and machine learning models were integrated to identify PD-related epigenetic regulatory genes, and clinical samples were used to verify their expression and immune regulatory functions. The key genes and pathways identified provide a potential molecular basis for early diagnosis and targeted therapy of PD. Methods Data Sources and Preprocessing The GSE7621 and GSE49036 datasets were downloaded from the Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/ ) as training sets. Data normalization was performed using the Sangerbox tool, followed by batch effect removal via the Combat algorithm to generate combined datasets. The principal component analysis (PCA) cluster plot was used to visualize the effects of inter-batch difference removal. Identification and Analysis of Differentially Expressed Genes (DEGs) To analyze gene expression differences between healthy controls (HC) and Parkinson’s disease (PD) patients in substantia nigra tissues, the "Limma" ( https://bioconductor.org/packages/release/bioc/html/limma.html ) package was applied to identify DEGs in the training sets using thresholds of |log2FoldChange| >1 and q-value < 0.05. The "EnhancedVolcano" R package was used to visualize DEGs in volcano plots. The top 50 significantly upregulated and downregulated genes in each sample were clustered based on log2FoldChange values, and heatmaps were generated using the "pheatmap" R package. Construction of Weighted Gene Co-expression Network (WGCNA) To identify PD-associated module genes, the "WGCNA" R package was used to construct a co-expression network for all samples in the training set. Hierarchical clustering analysis was performed to detect and remove outliers. A scale-free network was established, and genes were partitioned into modules. The correlation coefficient and p-value were calculated to analyze associations between modules and groups (PD vs. HC). Modules with the strongest correlations (|cor| >0.3 and p < 0.05) were selected, and genes within these modules were defined as PD-related module genes. Identification of Candidate Genes To identify PD-related genes associated with epigenetic factors, genes overlapping between DEGs, PD-related module genes, and epigenetic factor-related genes (from the EpiFactors database, v2.0) were extracted as candidate genes. Functional Enrichment Analysis The Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) [ 10 ] and Reactome enrichment analyses of DEGs were conducted using the ‘clusterProfiler’ ( https://bioconductor.org/packages/release/bioc/html/clusterProfiler.html ) package of R software. Findings of GO annotation analysis were categorized as follows: Biological Process (BP), Molecular Function (MF), and Cellular Components (CC)[1]. Significant enrichment was defined as adjusted P values < 0.05. Protein-Protein Interaction (PPI) Network Construction To establish the interplay of DEGs, PPI network analysis was performed using the STRING database ( https://cn.string-db.org/ ). A stringent interaction score threshold of 0.4 was employed to identify the most reliable and relevant interactions among the DEGs. Genes were denoted by nodes while connections between them were represented by edges. Subsequently, the main regulatory network was constructed and visualized using Cytoscape ( https://cytoscape.org/ ). The cytoHubba plugin of Cytoscape was used to identify the hub genes in the PPI network. Machine Learning To systematically evaluate the predictive capacity of candidate genes for PD, 90 machine learning models were constructed based on 12 key genes, incorporating algorithms such as XGBoost, SVM, QDA, KNN, and Lasso regularization. Model performance was assessed using metrics including area under the curve (AUC), accuracy, recall, and F-score. Feature importance weights were calculated across models to generate gene importance bar plots and average rank plots, identifying genes with stable contributions. Visualizations (e.g., AUC curves, precision plots, and importance rankings) were used to illustrate model performance and gene relevance. Validation Set Expression Verification and Diagnostic Model Evaluation Box/violin plots were drawn to identify the differentially expressed hub genes between HC and PD patients in the verification set. Then, the efficacy of distinguishing between HC and PD patients was tested by receiver operating characteristic (ROC) curve analysis. Gene Set Enrichment Analysis (GSEA) GSEA was performed on the training set using the "c2.cp.kegg.v7.0.symbols.gmt" reference gene set. Spearman correlation coefficients between key genes and genome-wide genes were calculated using the "psych" R package. Enriched KEGG pathways were identified using the "clusterProfiler" R package (significance thresholds: p < 0.05, FDR 1), and the top 10 significant pathways were visualized. Immune Cell Infiltration Analysis The R package ‘CIBERSORT’ ( https://github.com/Moonerss/CIBERSORT ) was applied to estimate immune cell infiltration abundance in PD and normal samples (excluding low-confidence samples with P > 0.05). Associations between key gene expression (ACTL6B, NAP1L2, etc.) and immune cell infiltration were analyzed by dividing samples into high/low expression groups based on median values. Wilcoxon or t-tests (threshold: P < 0.05) were used to compare immune cell abundance differences, visualized via boxplots. Reverse Transcription Quantitative Polymerase Chain Reaction (RT-qPCR) Clinical whole blood samples were collected from both HC and PD patients, using heparinized EDTA anticoagulant tubes. The blood samples were centrifuged twice at 3000 g for 10 min each at 4°C. Serum was then harvested and stored at -80°C until further analysis. Total RNA was extracted using TRIzol reagent (Thermo Fisher Scientific, Waltham, Massachusetts, USA). RNA concentration was measured using a Nano-100 instrument (Allsheng, Hangzhou, China), and cDNA synthesis was performed using the Hiscript II QRT Supermix for qPCR kit (Vazyme, Nanjing, China). The reaction conditions were 50°C for 15 minutes and 85°C for 5 seconds. The qPCR experiments were executed on an ABI7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) with the following cycling parameters: pre-denaturation at 95°C for 30 seconds, denaturation at 95°C for 10 seconds, annealing at 60°C for 30 seconds, 40 cycles. GAPDH was used as an internal control. Ct values were analyzed using the 2 -ΔΔCt method. The primer sequences were listed in Supplementary Table 1. Statistical analysis Bioinformatics statistical analyses were conducted using R language (version 4.0). All the data were processed by GraphPad Prism 7.0 statistical software (GraphPad, San Diego, CA, USA), and the data were presented as multiple groups of repeated data or means ± standard deviation. The comparisons between the two groups were analyzed by t-test. Statistical significance between two groups was determined by a t-test with a p-value < 0.05 considered significant. Results Identification of DEGs The GSE7621 and GSE49036 datasets were utilized as training sets, comprising 16 PD patient substantia nigra tissue samples and 9 HC substantia nigra tissue samples, as well as 15 PD patient (Braak stages 3–4 or 5–6) substantia nigra tissue samples and 8 HC substantia nigra tissue samples, respectively. Standardization was performed using the Sangerbox tool, followed by batch effect removal via Combat to generate combined datasets, thereby minimizing potential biases from batch effects in subsequent analyses. The integrated dataset consisted of 31 PD patient substantia nigra samples and 16 HC substantia nigra samples. Visualization of density distributions (Supplementary Fig. 1A), UMAP distributions (Supplementary Fig. 1B), boxplots (Supplementary Fig. 1C), and principal component analysis (Supplementary Fig. 1D) demonstrated differences in data distribution before and after batch correction, confirming effective elimination of batch effects. In the training set samples, 78 DEGs between hc and PD patients were identified using thresholds of |log2FC| >1 and q-value < 0.05, and visualized in a volcano plot (Fig. 1 A). Additionally, hierarchical clustering analysis was performed on the top 50 most significantly upregulated and downregulated genes in each sample, with the overall clustering results displayed in a heatmap (Fig. 1 B). Identification of PD-Related Module Genes and Key Genes Next, WGCNA analysis and module identification were performed on the training set samples to identify PD-related module genes. First, hierarchical clustering analysis was conducted on all samples to detect outliers, and a clustering dendrogram was generated to reveal clear clustering patterns between the PD and HC groups (Fig. 2 A). After removing outliers, the optimal soft threshold was determined using the pickSoftThreshold function to establish a scale-free network (Fig. 2 B). Gene clustering and module identification were achieved via the "dynamic tree cutting" algorithm, generating distinct co-expression modules (Fig. 2 C). Module-trait relationship analysis demonstrated significant correlations between specific modules and disease status (PD vs. HC). Two key modules exhibited the strongest associations: the blue module (cor = 0.37, p = 0.01) showed a positive correlation with PD, while the turquoise module (cor = − 0.43, p = 0.002) displayed a negative correlation (Fig. 2 D). Both modules met stringent thresholds (|cor| >0.3, p < 0.05), and genes within these modules were defined as PD-related module genes. The PD-related module genes identified above were merged with DEGs obtained from prior Limma differential analysis, yielding 550 PD-associated genes. Subsequently, 796 non-redundant epigenetic factor-related genes were extracted from the EpiFactors database. Intersection analysis between these PD-associated genes and epigenetic factor genes revealed 12 overlapping genes (Fig. 2 E). These overlapping genes, representing PD-related epigenetic regulators, were defined as key genes for subsequent analyses. Functional Enrichment and PPI Network of PD-Related Genes To systematically reveal the functional characteristics of PD-related genes, this study conducted multi-level functional annotation analysis on 550 PD-associated genes derived from the integration of DEGs and WGCNA. Functional enrichment analysis performed using the clusterProfiler package (p-value < 0.05) demonstrated that the gene sets were significantly enriched in BP such as "vesicle-mediated transport in synapse" and "regulated exocytosis" (Supplementary Fig. 2A). CC analysis indicated that key genes were primarily localized to subcellular structures including "synaptic membrane," "transport vesicle," and "transport vesicle membrane" (Supplementary Fig. 2B). MF analysis revealed significant associations with biomolecular properties such as "active monoatomic ion transmembrane transporter activity," "calmodulin binding," and "transmembrane transporter binding" (Supplementary Fig. 2C). KEGG enrichment analysis further validated the significant enrichment of these genes in pathways such as "Neurodegeneration-multiple diseases" and "Parkinson disease" (Supplementary Fig. 2D). Reactome pathway analysis supplemented these findings by highlighting critical molecular mechanisms, including "Transmission across Chemical Synapses" and "Neuronal System" (Supplementary Fig. 2E). Notably, visualization of pathway interaction networks constructed via Cytoscape demonstrated that these enriched pathways formed distinct functional network modules through shared genes, with "calcium-ion regulated exocytosis" and "inorganic cation transmembrane transporter activity" acting as core hub nodes (Supplementary Fig. 3). This systematic multi-dimensional functional annotation provides novel molecular network insights into PD pathogenesis. Furthermore, to explore protein-protein interactions encoded by PD-related genes, a PPI network was constructed using the STRING database and visualized with Cytoscape (Supplementary Fig. 4). Machine Learning Evaluation of Key Genes This study systematically evaluated the predictive efficacy of 12 key genes for PD by constructing a multidimensional evaluation framework based on 90 machine learning models. In terms of model performance, the AUC curves of the top 50 models (Fig. 3 A) revealed that the KNN algorithm (k = 1/2) exhibited strong discriminatory power (AUC > 0.7), while the XGBoost + Lasso combined model demonstrated stable performance in cross-gene prediction (average precision: 66%) (Fig. 3 B). Through Recall-Fscore dual-index analysis, QDA and multiple SVM variants (including SVM-default, SVM-CV:5, etc.) achieved optimal balance between model recall and prediction accuracy (Fig. 3 C-D). Gene importance analysis highlighted significant algorithm-dependent characteristics. By integrating feature weights and average ranks across models, we identified 10 core predictive genes: EID2B, PRMT8, ACTL6B, SUDS3, SMYD3, EID2, YWHAZ, UBE2T, CDK5, and NAP1L2. Notably, EID2B was consistently recognized as a critical feature across all models, suggesting its stable predictive value transcending algorithms (Fig. 4 A-B). The integrated model performance metrics and gene importance analysis provide critical insights for further research. Validation of Key Gene Expression To systematically validate the reliability of key genes in the substantia nigra of PD patients, this study analyzed two independent cohorts from the GEO database. Wilcoxon rank-sum tests were performed to compare the expression of key genes in the validation sets: GSE20141 (10 PD patient SNpc neuron samples and 8 HC SNpc neuron samples) and GSE42966 (9 PD patient substantia nigra tissue samples and 6 HC samples) (Fig. 5 A-B). Results showed discrepant differential expression patterns for PRMT8, SMYD3, and UBE2T between the two validation sets. Additionally, VAV1 expression data were absent in GSE42966. ROC curve analysis evaluated the predictive performance of key genes. HSPA1A exhibited weak disease prediction capability (AUC 0.7, three genes—ACTL6B, NAP1L2, and PRMT8—were identified as key epigenetics-related genes in the PD substantia nigra (Fig. 5 D), while BABAM1 and EID2 were highlighted as key genes associated with SNpc neurons and epigenetic regulation (Fig. 5 C). To further confirm bioinformatics findings, blood samples from PD patients (PD group) and healthy controls (HC group) were analyzed via RT-qPCR. The results showed that the mRNA expression levels of ACTL6B, NAP1L2, PRMT8, BABAM1, and EID2 were significantly decreased in the PD group (Fig. 6 ). GSEA Enrichment Analysis To elucidate the potential biological functions of key genes in PD pathogenesis, gene set enrichment analysis was performed using the KEGG database. Pathways meeting significance thresholds (p < 0.05, FDR 1) were prioritized. The top 10 enriched pathways (Fig. 7 ) suggested that ACTL6B, NAP1L2, PRMT8, BABAM1, and EID2 may contribute to disease progression by regulating cellular metabolic reprogramming, proteasome activity, and related processes. These findings offer critical clues for deciphering molecular mechanisms. Immune Infiltration Analysis To explore immune cell-PD interactions, CIBERSORT was applied to assess the infiltration abundance of 22 immune cell types in training set transcriptomic data (low-confidence samples with algorithm P > 0.05 were excluded). Results revealed significant differences in CD8 + T cells, mast cells, and macrophage subtypes (M1/M2) across samples (Fig. 8 A), indicating immune microenvironment reprogramming in PD. Further analysis demonstrated significant correlations between key gene expression (ACTL6B, NAP1L2, PRMT8, BABAM1, EID2) and specific immune cell infiltration. After stratifying samples into high/low expression groups by median values, Wilcoxon rank-sum and independent t-tests showed that high ACTL6B expression correlated with elevated CD8 + T cell and M2 macrophage infiltration, whereas PRMT8 and EID2 were linked to negative regulation of resting memory CD4 + T cells (Fig. 8 B-F). Discussion Biomarker research for PD is currently in a rapidly developing phase, providing assistance for critical clinical needs such as early diagnosis, progression monitoring, and therapeutic targets [ 11 ] . This study employed multi-dimensional integrated approaches including differential gene screening, Weighted Gene Co-expression Network Analysis (WGCNA), functional enrichment analysis, machine learning, Gene Set Enrichment Analysis (GSEA), and immune infiltration analysis, combined with external datasets, to screen key differentially expressed genes in PD. Partial gene expression was validated using RT-qPCR. This study found that epigenetic genes such as ACTL6B, PRMT8, NAP1L2, BABAM1 and EID2 were significantly related to the pathological process of PD. Key genes can regulate synaptic transmission, cellular metabolic reprogramming, and proteasome activity in multiple dimensions, providing new molecular network insights into the pathogenesis of PD. In addition, immunoinfiltration analysis showed changes in the abundance of immune cells in patients, indicating that the expression of key genes was significantly correlated with specific immune cell infiltration. ACTL6B belongs to the actin-associated protein family (ACTL6), which is involved in the chromatin remodeling complex [ 12 ] , regulates nucleosomal composition and gene transcription, and plays a key role in neuronal development and synaptic plasticity [ 13 ] . At present, the direct role of ACTL6B in Parkinson's disease has not been reported, but members of the chromatin remodeling complex (e.g. KAT8/KANSL1) regulate α-synuclein (SNCA) expression in PD through histone acetylation, thereby regulating parkinson-related genes [ 14 ] . It suggested that ACTL6B may indirectly affect neuronal survival by regulating neuronal differentiation and synaptic plasticity. NAP1L2 is a nucleosome assembly protein, which is involved in the regulation of chromatin structure and gene expression as a histone chaperone [ 15 ] . Recent studies have found no direct link between NAP1L2 and PD, but found that its mRNA is regulated by m6A methylation, and m6A modifications (such as METTL3/ MeTTL14-mediated NAP1L2 methylation) are significantly dysregulated in PD, which may affect disease progression through regulation of neuronal gene expression [ 16 ] . PRMT8, a member of the protein arginine methyltransferase family (PRMTs), is an important epigenetic regulatory enzyme [ 17 ] . PRMT8 plays a key role in shaping the morphology of dendritic spines by catalyzing the methylation of arginine residues and regulating the activity of target proteins [ 18 ] . Synaptic dysfunction is one of the early pathological features of Parkinson's disease, so PRMT8 may indirectly affect the survival and function of dopamine neurons by regulating synaptic structural stability. BABAM1 is a member of the BRISC/BRCA1-A complex, which is involved in cell cycle regulation and inflammatory signaling pathways [ 19 ] . At present, the study does not mention the association or specific role of BABAM1 and PD. However, the BRISC complex is associated with neuroinflammation, which is an important pathological feature of PD [ 20 ] . Combined with the results of this study, it is speculated that it may affect PD progression through epigenetic-immune interaction network. EID2 belongs to the EID gene family (E1A-like inhibitor of differentiation), which is usually involved in transcriptional regulation and cell differentiation [ 21 ] . At present, there are few direct studies on EID2. In focal injury of articular cartilage, EID2 expression is significantly higher than that in uninjured tissue, and its up-regulation may be related to inflammatory response, hypoxia or cellular stress [ 22 ] . Therefore, EID2 may be involved in the regulation of Parkinson's immune microenvironment by influencing the differentiation and function of immune cells. Through GO and KEGG enrichment analysis, it was found that PD-related genes were significantly enriched in "vesicle-mediated transport in synapse", "calmodulin binding" and "transmembrane transporter". GSEA enrichment analysis showed that ACTL6B, NAP1L2, PRMT8, BABAM1, EID2 and other key genes were involved in significant pathways including cell metabolic reprogramming and regulation of proteasome activity. According to the study of ManWK et al. [ 23 ] , the overexpression of α-Synuclein can interfere with the speed of vesicle refilling, destroy the circulation of synaptic vesicles, and lead to a decrease in the release of dopamine and other transmitters. Stavsky A et al. [ 24 ] found that the decreased mitochondrial calcium buffering capacity in PD led to the accumulation of reactive oxygen species and energy depletion, exacerbating abnormal synaptic transmission. These pathways are consistent with our findings. These suggest that PD-related genes play an important role in neurotransmitter release, synaptic transmission and cell metabolic. In this study, it was found that the infiltration abundance of immune cells such as CD8 + T cells, mast cells and macrophage subtypes (M1/M2) in PD patients was significantly different. The expression of key genes is related to the infiltration of specific immune cells. In PD-related studies, the abnormality of immune cells and their association with genes have been reported many times. PD risk genes (such as LRRK2 and GBA) are highly expressed in monocytes and microglia and may affect disease risk by regulating inflammatory response [ 25 ] . The abnormal DNA methylation pattern of peripheral immune cells (such as CD14 + monocytes and CD4 + T cells) in PD patients may promote inflammation by regulating gene expression [ 26 ] . The key genes in this study were less involved in previous studies. This study found the correlation between key genes and immune cell infiltration, providing new clues for understanding the immunopathological mechanism of PD. This study is mainly based on bioinformatics analysis and clinical sample validation, and the specific molecular mechanisms of key genes regulating immune cell infiltration and pathway activity have not been clarified through cell or animal experiments. Although further mechanism verification is needed, this study lays a theoretical foundation for early diagnosis of PD and targeted epigenetic regulation to reshape the immune microenvironment. In the future, we will focus on the functional analysis of key genes, further elucidate their causal relationship in the occurrence and development of PD, and promote the transformation of basic research into clinical applications. Conclusion This study adopted integrated multi-omics and machine learning methods, combined with differential gene screening, WGCNA, GSEA and immune infiltration analysis, and verified by RT-qPCR, successfully identified epigenetic and immune hub genes associated with PD, revealing their links with immune microenvironment and key pathways, and laying a theoretical foundation for early diagnosis and precise treatment of PD. Declarations Correspondence Guiyue Meng, the Second Hospital of Shandong University, Jinan, Shandong, China. E-mail: [email protected] . Ethics approval and consent to participate The study was approved by the Ethics Committee of The Second Hospital of Shandong University. All principles in the Declaration of Helsinki were strictly followed in this study. The specimens used for RT-qPCR validation were obtained from inpatients. All patients provided signed informed consent forms, authorizing further research on blood specimens. All cited statistics from published studies and publicly available database were approved by a relevant review board. Clinical trial registration Not applicable. This study did not involve clinical trials or patient interventions, as it primarily analyzed publicly available transcriptomic datasets and institutional biobank samples with informed consent. Consent for publication The data used in this article are from publicly available databases, and the source databases have obtained informed consent to publish the researches and any accompanying images without disclosing personal information. No personal information was disclosed in this study. Competing interests The authors declare no competing interests. Authors information Author List: Guiyue Meng * 1, , Yunqi Sun 1, . Affiliate Institutions: Funding None. Author Contribution Guiyue Meng analyzed the data and wrote the manuscript. Yunqi Sun edited the manuscript and made suggestions for revision. All authors declare no disputes of interest. Acknowledgements Our sincere thanks to all participants included in this study and developers of the R software and all the related algorithms and websites. Data Availability The GSE7621 and GSE49036 datasets were downloaded from the Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/) as training sets. References Durmaz Celik N, Ozben S, Ozben T. Unveiling Parkinson's disease through biomarker research: current insights and future prospects. Crit Rev Clin Lab Sci. 2024;61(7):529–45. 10.1080/10408363.2024.2331471 . Zhang D, Zhang J, Wang Y, et al. Targeting epigenetic modifications in Parkinson's disease therapy. Med Res Rev. 2023;43(5):1748–77. 10.1002/med.21962 . Mourtzi T, Kazanis I. Endogenous versus exogenous cell replacement for Parkinson's disease: where are we at and where are we going? Neural Regen Res. 2022;17(12):2637–42. 10.4103/1673-5374.336137 . Cao Y, Xu Y, Cao M, et al. Fluid-based biomarkers for neurodegenerative diseases. Ageing Res Rev Published online March. 2025;21. 10.1016/j.arr.2025.102739 . Zhang D, Zhang J, Wang Y, et al. Targeting epigenetic modifications in Parkinson's disease therapy. Med Res Rev. 2023;43(5):1748–77. 10.1002/med.21962 . Guhathakurta S, Kim J, Adams L, et al. Targeted attenuation of elevated histone marks at SNCA alleviates α-synuclein in Parkinson's disease. EMBO Mol Med. 2021;13(2):e12188. 10.15252/emmm.202012188 . Zhang Z, Wang R, Zhou H et al. Inhibition of EHMT1/2 rescues synaptic damage and motor impairment in a PD mouse model. Cell Mol Life Sci. 2024;81(1):128. Published 2024 Mar 12. 10.1007/s00018-024-05176-5 de Mena L, Parés G, Garrido A, et al. α-Synuclein Gene Hypomethylation in LRRK2 Parkinson's Disease Patients. Mov Disord. 2025;40(3):550–5. 10.1002/mds.30094 . Liu G, Ni C, Zhan J, et al. Mitochondrial haplogroups and cognitive progression in Parkinson's disease. Brain. 2023;146(1):42–9. 10.1093/brain/awac327 . Kanehisa M, Furumichi M, Sato Y, Matsuura Y, Ishiguro-Watanabe M. KEGG: biological systems database as a model of the real world. Nucleic Acids Res. 2025;53(D1):D672–7. 10.1093/nar/gkae909 . Zhang L, Liu Y, Lu Y, Wang G. Targeting epigenetics as a promising therapeutic strategy for treatment of neurodegenerative diseases. Biochem Pharmacol. 2022;206:115295. 10.1016/j.bcp.2022.115295 . Bell S, Rousseau J, Peng H, et al. Mutations in ACTL6B Cause Neurodevelopmental Deficits and Epilepsy and Lead to Loss of Dendrites in Human Neurons. Am J Hum Genet. 2019;104(5):815–34. 10.1016/j.ajhg.2019.03.022 . Cali E, Quirin T, Rocca C, et al. Clinical and genetic delineation of autosomal recessive and dominant ACTL6B-related developmental brain disorders. Genet Med Published online September. 2024;17. 10.1016/j.gim.2024.101251 . Hicks AR, Reynolds RH, O'Callaghan B, et al. The non-specific lethal complex regulates genes and pathways genetically linked to Parkinson's disease. Brain. 2023;146(12):4974–87. 10.1093/brain/awad246 . Hu M, Xing L, Zhang L, et al. NAP1L2 drives mesenchymal stem cell senescence and suppresses osteogenic differentiation. Aging Cell. 2022;21(2):e13551. 10.1111/acel.13551 . Yan J, Wang Z, Li Y, Li R, Xiang K. m6A-related genes and their role in Parkinson's disease: Insights from machine learning and consensus clustering. Med (Baltim). 2024;103(45):e40484. 10.1097/MD.0000000000040484 . Hsu SH, Hung WC. Protein arginine methyltransferase 3: A crucial regulator in metabolic reprogramming and gene expression in cancers. Cancer Lett. 2023;554:216008. 10.1016/j.canlet.2022.216008 . Lo LH, Dong R, Lyu Q, Lai KO. The Protein Arginine Methyltransferase PRMT8 and Substrate G3BP1 Control Rac1-PAK1 Signaling and Actin Cytoskeleton for Dendritic Spine Maturation. Cell Rep. 2020;31(10):107744. 10.1016/j.celrep.2020.107744 . Jin F, Zhu Y, Liu M, et al. Babam2 negatively regulates osteoclastogenesis by interacting with Hey1 to inhibit Nfatc1 transcription. Int J Biol Sci. 2022;18(11):4482–96. 10.7150/ijbs.72487 . Published 2022 Jul 11. Zhang W, Liu K, Ren GM, et al. BRISC is required for optimal activation of NF-κB in Kupffer cells induced by LPS and contributes to acute liver injury. Cell Death Dis. 2023;14(11):743. 10.1038/s41419-023-06268-z . Published 2023 Nov 15. Fu X, Luo L, Yi R, et al. Transcriptome profiling in Eid1-KO mice brain shows that Eid1 links cell proliferation in the brain. Gene. 2019;717:143998. 10.1016/j.gene.2019.143998 . Aşık MD, Gürsoy S, Akkaya M, Kozacı LD, Doğan M, Bozkurt M. Microarray analysis of cartilage: comparison between damaged and non-weight-bearing healthy cartilage. Connect Tissue Res. 2020;61(5):456–64. 10.1080/03008207.2019.1611797 . Man WK, Tahirbegi B, Vrettas MD, et al. The docking of synaptic vesicles on the presynaptic membrane induced by α-synuclein is modulated by lipid composition. Nat Commun. 2021;12(1):927. 10.1038/s41467-021-21027-4 . Published 2021 Feb 10. Stavsky A, Stoler O, Kostic M et al. Aberrant activity of mitochondrial NCLX is linked to impaired synaptic transmission and is associated with mental retardation [published correction appears in Commun Biol. 2021;4(1):755. doi: 10.1038/s42003-021-02312-w.]. Commun Biol. 2021;4(1):666. Published 2021 Jun 2. 10.1038/s42003-021-02114-0 Navarro E, Udine E, Lopes KP, et al. Dysregulation of mitochondrial and proteolysosomal genes in Parkinson's disease myeloid cells. Nat Aging. 2021;1(9):850–63. 10.1038/s43587-021-00110-x . Andersen MS, Leikfoss IS, Brorson IS et al. Epigenome-wide association study of peripheral immune cell populations in Parkinson's disease. NPJ Parkinsons Dis. 2023;9(1):149. Published 2023 Oct 31. 10.1038/s41531-023-00594-x Additional Declarations No competing interests reported. Supplementary Files FigS1.jpg Supplementary Figure 1. The datasets were normalized, and the batch effects removed. A. Comparison of density distributions before and after batch effect removal. B. Comparison of UMAP distributions before and after batch effect correction. C. Comparison of data distributions before and after batch effect adjustment. D. Principal component analysis (PCA) analysis illustrating sample variations across disease groups. Left: Two-dimensional PCA analysis; Right: Three-dimensional PCA analysis. FigS2.jpg Supplementary Figure 2. Functional enrichment of PD-related genes. FigS3.jpg Supplementary Figure 3. Network of enriched pathways for candidate genes. FigS4.jpg Supplementary Figure 4. Protein-Protein Interaction Network of PD-related genes. SupplementaryTable1.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7376065","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":517850347,"identity":"3666bb68-2dcd-4091-be6f-deb74ce66ed3","order_by":0,"name":"Guiyue Meng","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAz0lEQVRIiWNgGAWjYPACCSBmPnDgww/StLAlHpzZQ5pNPMaHOdiIUKfb3nv45dc2i8QNx3s+HGbgYZDnFzuAX4vZmXNp1rJtEokbzpzdcLjAgsFw5uwEAlpu5JgZS4K03MjdcHgGD0OCwW1CWu6/gWq5/+bBYR42YrTc4DF++BFsCw8DkVrO5JgxM5yTMJ55Js0AGMgSRPjl+Bnjjz/K6mT7jh9+/OHDDxt5fmkCWoCATZqHgcGxAcKRIKgcBJg/ApOJPVFKR8EoGAWjYGQCAK4FTHQRzzMtAAAAAElFTkSuQmCC","orcid":"","institution":"Second Hospital of Shandong University","correspondingAuthor":true,"prefix":"","firstName":"Guiyue","middleName":"","lastName":"Meng","suffix":""},{"id":517850350,"identity":"480156bc-cdd8-442d-80fd-26b9e4f1422a","order_by":1,"name":"Yunqi Sun","email":"","orcid":"","institution":"Second Hospital of Shandong University","correspondingAuthor":false,"prefix":"","firstName":"Yunqi","middleName":"","lastName":"Sun","suffix":""}],"badges":[],"createdAt":"2025-08-14 17:38:10","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7376065/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7376065/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":92126103,"identity":"3c004319-c0c2-4e86-a001-7f682cbc2e19","added_by":"auto","created_at":"2025-09-25 01:32:21","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":51833,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscriptend.docx","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/502254c7055ad8463c9b3ca9.docx"},{"id":92127016,"identity":"0aa6cc81-2f17-4249-94e3-50dc9fb5970c","added_by":"auto","created_at":"2025-09-25 01:40:21","extension":"json","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4907,"visible":true,"origin":"","legend":"","description":"","filename":"96cc48ecc73d4a1287a0ba9fdcfeada4.json","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/08481ae47e79bc38aafaf75d.json"},{"id":92126114,"identity":"47404534-22b1-4834-9c4e-609c9e7836bf","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"xml","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":93716,"visible":true,"origin":"","legend":"","description":"","filename":"96cc48ecc73d4a1287a0ba9fdcfeada41enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/e631f9f8ef09dcb08bbd426f.xml"},{"id":92126115,"identity":"d514e791-c7f4-4d27-a6bd-730dbe5b2c44","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"jpg","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2234523,"visible":true,"origin":"","legend":"","description":"","filename":"Fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/8a5743280253bbbb72f5aacd.jpg"},{"id":92126123,"identity":"59b21661-bfd4-41a0-a60b-6074d2e97d76","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"jpg","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2728160,"visible":true,"origin":"","legend":"","description":"","filename":"Fig2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/832530c1cae921d02bc88843.jpg"},{"id":92126130,"identity":"09eb8ddb-da74-4c63-86f0-78572061794c","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"jpg","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8351027,"visible":true,"origin":"","legend":"","description":"","filename":"Fig3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/1283d587927c53684eba2f2f.jpg"},{"id":92127187,"identity":"7bf56e7b-a91c-4bea-945d-ac5aabbdb965","added_by":"auto","created_at":"2025-09-25 01:48:22","extension":"jpg","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2787893,"visible":true,"origin":"","legend":"","description":"","filename":"Fig4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/16948c83d1fa73a42a5f6b5c.jpg"},{"id":92127025,"identity":"ad403674-fe04-4891-ba86-3ee4105524b7","added_by":"auto","created_at":"2025-09-25 01:40:22","extension":"jpg","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3900683,"visible":true,"origin":"","legend":"","description":"","filename":"Fig5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/491bc9a2eb536c273653324f.jpg"},{"id":92126120,"identity":"bb80c98c-bc31-431e-be1f-22e151cb17e9","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"jpg","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":524715,"visible":true,"origin":"","legend":"","description":"","filename":"Fig6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/e817d1d36dae605f2b20f3ed.jpg"},{"id":92127020,"identity":"46c4e84a-3a1d-44b9-b430-4438af6432d6","added_by":"auto","created_at":"2025-09-25 01:40:22","extension":"jpg","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4356537,"visible":true,"origin":"","legend":"","description":"","filename":"Fig7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/45d09ef625fc691132b42ea0.jpg"},{"id":92126140,"identity":"b9ec71d2-d616-4514-bf7f-d2dc35d39e35","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"jpg","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4258996,"visible":true,"origin":"","legend":"","description":"","filename":"Fig8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/f52eddfeea5380d530c7e516.jpg"},{"id":92127026,"identity":"6b80a56c-7860-48cb-a242-65cb29a5ea7c","added_by":"auto","created_at":"2025-09-25 01:40:22","extension":"jpg","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3275064,"visible":true,"origin":"","legend":"","description":"","filename":"FigS1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/3dd7c3f7b5b96c9f734b03f7.jpg"},{"id":92127186,"identity":"42f23302-4c13-414e-8cf2-4266ab487fd3","added_by":"auto","created_at":"2025-09-25 01:48:22","extension":"jpg","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3237690,"visible":true,"origin":"","legend":"","description":"","filename":"FigS2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/6339941698afb909c2c20e72.jpg"},{"id":92126134,"identity":"db62fff5-afce-4205-83f9-051b69ee28c1","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"jpg","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1681155,"visible":true,"origin":"","legend":"","description":"","filename":"FigS3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/74fc51c9001a3ffe5086dbce.jpg"},{"id":92127023,"identity":"0632ed47-50c1-43b5-a1b5-4e2d7b06d0e5","added_by":"auto","created_at":"2025-09-25 01:40:22","extension":"jpg","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2533265,"visible":true,"origin":"","legend":"","description":"","filename":"FigS4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/42760d875e7357c0892caf02.jpg"},{"id":92126133,"identity":"5b9bf023-1087-4da6-967b-a6d9bebca38c","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"png","order_by":27,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":976550,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig1.png","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/282428380a9dda9a086524c7.png"},{"id":92126136,"identity":"f6a28be4-5cb5-4e78-801a-e98b7a5ae9af","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"png","order_by":28,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":832547,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig2.png","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/60bcc3d6208dfa696a73ec18.png"},{"id":92127022,"identity":"12fb1750-fe69-4316-9d19-bba83126601d","added_by":"auto","created_at":"2025-09-25 01:40:22","extension":"png","order_by":29,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3722827,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig3.png","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/551afd275299f1ff94695fe7.png"},{"id":92126139,"identity":"cfc1fc04-44dc-4a24-ab70-cabc073284bf","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"png","order_by":30,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":697789,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig4.png","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/40ffceb35a19b37538f21b32.png"},{"id":92126127,"identity":"718558bc-9932-4264-a855-73c7e89aacfb","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"png","order_by":31,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2016978,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig5.png","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/95938fc5ed4fd99c5d4f7e01.png"},{"id":92127028,"identity":"a254b84a-1c82-4932-9d40-d514d8ab177e","added_by":"auto","created_at":"2025-09-25 01:40:22","extension":"png","order_by":32,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":188129,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig6.png","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/5f700783fa83b2a3f90fe3ec.png"},{"id":92126143,"identity":"edfa95cb-d9f7-42ee-9698-9d85fdd3607c","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"png","order_by":33,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2591821,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig7.png","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/8f4f962e26e85c61349a9320.png"},{"id":92127029,"identity":"26b2b2c9-9575-4d65-853a-1e56b6a45966","added_by":"auto","created_at":"2025-09-25 01:40:22","extension":"png","order_by":34,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1403560,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig8.png","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/4e93854d0e6d1cf845916201.png"},{"id":92126125,"identity":"433b2f40-984b-4650-a217-5a2357cd7379","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"png","order_by":35,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1409467,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigS1.png","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/17800eeedf0dd6813e287d9a.png"},{"id":92126132,"identity":"d2907fa5-9e27-4bb0-9a23-cdee76109682","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"png","order_by":36,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":885682,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigS2.png","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/962cdc40ca95c9d7086eee88.png"},{"id":92126131,"identity":"941fd744-4857-4140-a0c3-8f765c42ee91","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"png","order_by":37,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1074503,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigS3.png","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/804be75687510fa8c8d51526.png"},{"id":92127027,"identity":"21cbd68a-f6e3-4f93-a7ec-bc3422ae92d6","added_by":"auto","created_at":"2025-09-25 01:40:22","extension":"png","order_by":38,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2225101,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigS4.png","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/033379084669baf8b131dfd5.png"},{"id":92126144,"identity":"f888c0cd-a179-4863-9112-a96047f92384","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"xml","order_by":39,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":89462,"visible":true,"origin":"","legend":"","description":"","filename":"96cc48ecc73d4a1287a0ba9fdcfeada41structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/9a0ac6e8beacc500fabf7949.xml"},{"id":92126142,"identity":"915c9475-84ab-469c-969e-f45e9e177e6b","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"html","order_by":40,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":101438,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/0329a13312901eb106cf95d9.html"},{"id":92126105,"identity":"03626d59-8ac4-44e8-9e96-d5bf8b8881c8","added_by":"auto","created_at":"2025-09-25 01:32:21","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2234523,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification of differentially expressed genes (DEGs).\u003c/p\u003e\n\u003cp\u003eA. Volcano Plot of DEGs. Genes with significant differential expression are represented by red dots (upregulated) and blue dots (downregulated), while genes without significant differential expression are represented by gray dots. The x-axis represents the fold change in gene expression between different groups, and the y-axis represents the statistical significance of gene expression changes.\u003c/p\u003e\n\u003cp\u003eB. Heatmap: Clustering of the top 50 upregulated/downregulated genes per sample.\u003c/p\u003e","description":"","filename":"Fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/31736967305f5290afbd8a33.jpg"},{"id":92126117,"identity":"571b0702-3791-4278-ad87-ab3d945fe693","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2728160,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification of Parkinson’s disease (PD)-related module genes and key genes.\u003c/p\u003e\n\u003cp\u003eA. Hierarchical clustering of training set samples.\u003c/p\u003e\n\u003cp\u003eB. Determination of soft threshold for scale-free networks.\u003c/p\u003e\n\u003cp\u003eC. Co-expression module clustering.\u003c/p\u003e\n\u003cp\u003eD. Heatmap of module-group correlations.\u003c/p\u003e\n\u003cp\u003eE. Venn diagram of overlapping PD-related and epigenetic factor genes.\u003c/p\u003e","description":"","filename":"Fig2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/df73471ac66479a144b9e596.jpg"},{"id":92126107,"identity":"768bdb14-e304-48fc-a15b-6dc3a86bb409","added_by":"auto","created_at":"2025-09-25 01:32:21","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":8351027,"visible":true,"origin":"","legend":"\u003cp\u003eMachine learning for key genes.\u003c/p\u003e\n\u003cp\u003eA. Area under the curve (AUC) of top 50 models.\u003c/p\u003e\n\u003cp\u003eB. Accuracy of top 50 models.\u003c/p\u003e\n\u003cp\u003eC. Recall of top 50 models.\u003c/p\u003e\n\u003cp\u003eD. F-score of top 50 models.\u003c/p\u003e","description":"","filename":"Fig3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/4a8c7180f4895012bf98e70b.jpg"},{"id":92126106,"identity":"412f780c-1333-49fe-a792-1b05415e835a","added_by":"auto","created_at":"2025-09-25 01:32:21","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2787893,"visible":true,"origin":"","legend":"\u003cp\u003eGene importance analysis.\u003c/p\u003e\n\u003cp\u003eA. Feature importance bar plots across models.\u003c/p\u003e\n\u003cp\u003eB. Average gene importance ranking.\u003c/p\u003e","description":"","filename":"Fig4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/a81bec38ef08daac78ce28e3.jpg"},{"id":92126109,"identity":"6bdb4150-84b1-44ff-bc7f-cad7b14b87f6","added_by":"auto","created_at":"2025-09-25 01:32:21","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":3900683,"visible":true,"origin":"","legend":"\u003cp\u003eValidation of key gene expression.\u003c/p\u003e\n\u003cp\u003eA. Box plots and violin plots illustrating expression differences of key genes in the GSE20141 dataset.\u003c/p\u003e\n\u003cp\u003eB. Box plots and violin plots demonstrating expression variations of key genes in the GSE42966 dataset.\u003c/p\u003e\n\u003cp\u003eC. ROC curves evaluating the diagnostic performance of key genes in the GSE20141 dataset.\u003c/p\u003e\n\u003cp\u003eD. ROC curves assessing the predictive capability of key genes in the GSE42966 dataset.\u003c/p\u003e","description":"","filename":"Fig5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/f0941e4401c5e64f90b31cf3.jpg"},{"id":92126113,"identity":"1a54e7a4-05ef-4d39-a52c-f20e528389ec","added_by":"auto","created_at":"2025-09-25 01:32:21","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":524715,"visible":true,"origin":"","legend":"\u003cp\u003eReverse Transcription Quantitative Polymerase Chain Reaction (RT-qPCR) validation of key genes in PD and healthy controls (HC) blood samples.\u003csup\u003e *\u003c/sup\u003eP\u0026lt;0.05\u003csup\u003e**\u003c/sup\u003eP\u0026lt;0.01 \u003cem\u003evs.\u003c/em\u003e HC Group.\u003c/p\u003e","description":"","filename":"Fig6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/3b43115ba82c0c00adeb78ba.jpg"},{"id":92126121,"identity":"b4b05ea5-b273-4654-a1c7-fada7a7742ad","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":4356537,"visible":true,"origin":"","legend":"\u003cp\u003eGene set enrichment analysis (GSEA). A-E. GSEA results for ACTL6B, NAP1L2, PRMT8, BABAM1, and EID2.\u003c/p\u003e","description":"","filename":"Fig7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/246315215e6f7d9d6cc056bc.jpg"},{"id":92127021,"identity":"f880c19a-fed3-4ab1-9133-09fd92a83317","added_by":"auto","created_at":"2025-09-25 01:40:22","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":4258996,"visible":true,"origin":"","legend":"\u003cp\u003eImmune infiltration analysis.\u003c/p\u003e\n\u003cp\u003eA. Proportion of immune-infiltrating cells in training set samples.\u003c/p\u003e\n\u003cp\u003eB. Correlation analysis between the hub gene ACTL6B and differential immune cells. C. Correlation analysis between the hub gene NAP1L2 and differential immune cells. D. Correlation analysis between the hub gene PRMT8 and differential immune cells. E. Correlation analysis between the hub gene BABAM1 and differential immune cells. F. Correlation analysis between the hub gene EID2 and differential immune cells.\u003c/p\u003e","description":"","filename":"Fig8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/effd5615b0b4068f3b0bfd14.jpg"},{"id":93111039,"identity":"9c5728c4-5467-4b07-be5f-fe5715dadc25","added_by":"auto","created_at":"2025-10-09 07:48:32","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":30030822,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/d95dc48b-d240-4bb8-8c7d-05b34d4a47ce.pdf"},{"id":92126116,"identity":"be96b9cd-25db-4124-98c9-47ded22fd405","added_by":"auto","created_at":"2025-09-25 01:32:22","extension":"jpg","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":3275064,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Figure 1. \u003c/strong\u003eThe datasets were normalized, and the batch effects removed.\u003c/p\u003e\n\u003cp\u003eA. Comparison of density distributions before and after batch effect removal.\u003c/p\u003e\n\u003cp\u003eB. Comparison of UMAP distributions before and after batch effect correction.\u003c/p\u003e\n\u003cp\u003eC. Comparison of data distributions before and after batch effect adjustment.\u003c/p\u003e\n\u003cp\u003eD. Principal component analysis (PCA) analysis illustrating sample variations across disease groups. Left: Two-dimensional PCA analysis; Right: Three-dimensional PCA analysis.\u003c/p\u003e","description":"","filename":"FigS1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/e0ee640cbfa45e78366d4255.jpg"},{"id":92126111,"identity":"207a5638-4217-48fe-a471-9a2609adb2bf","added_by":"auto","created_at":"2025-09-25 01:32:21","extension":"jpg","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":3237690,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Figure 2. \u003c/strong\u003eFunctional enrichment of PD-related genes.\u003c/p\u003e","description":"","filename":"FigS2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/b5d4984d69c2982da4b37412.jpg"},{"id":92127017,"identity":"3af6e6af-1bdc-4d6c-b7e3-81eb5f4fe323","added_by":"auto","created_at":"2025-09-25 01:40:21","extension":"jpg","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":1681155,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Figure 3. \u003c/strong\u003eNetwork of enriched pathways for candidate genes.\u003c/p\u003e","description":"","filename":"FigS3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/48d0f602c095a92ccae802f5.jpg"},{"id":92127018,"identity":"19b5230f-5a3e-4c2e-8bb0-b77e1cb03822","added_by":"auto","created_at":"2025-09-25 01:40:22","extension":"jpg","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":2533265,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Figure 4. \u003c/strong\u003eProtein-Protein Interaction Network of PD-related genes.\u003c/p\u003e","description":"","filename":"FigS4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/9ee30ce14678b0b4123beb8a.jpg"},{"id":92127015,"identity":"fa16e328-8c4a-4006-90b5-342b20b78332","added_by":"auto","created_at":"2025-09-25 01:40:21","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":15146,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable1.docx","url":"https://assets-eu.researchsquare.com/files/rs-7376065/v1/9ba2e6cf6d5539fac9423c6b.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Correlation analysis and clinical validation of Parkinson's disease and epigenetic factor-related genes based on transcriptome data","fulltext":[{"header":"Introduction","content":"\u003cp\u003eParkinson's disease (PD) is a chronic and progressive neurodegenerative disorder, mainly manifested by the selective loss of dopaminergic neurons in the substantia nigra pars compacta and the abnormal aggregation of α-synuclein to form Lewy bodies\u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]\u003c/sup\u003e, resulting in core motor symptoms such as bradykinesia, resting tremor, and muscle rigidity. Typical neurodegenerative features of PD include progressive apoptosis of dopaminergic neurons, synaptic dysfunction, and a widespread neuroinflammatory response\u003csup\u003e[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]\u003c/sup\u003e. Current clinical diagnosis is mainly dependent on the appearance of motor symptoms, at which time about 50%-70% of dopaminergic neurons have been los\u003csup\u003et[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]\u003c/sup\u003e. Existing therapies lack drugs that target core pathological mechanisms such as epigenetic inheritance or protein misfolding. The discovery of reliable biomarkers can improve patient outcomes by breaking bottlenecks in early diagnosis and accelerating the development of targeted therapies\u003csup\u003e[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eEpigenetics regulates gene expression through mechanisms such as DNA methylation, histone modification and non-coding RNA, affecting cell function and phenotype\u003csup\u003e[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]\u003c/sup\u003e, and is an important bridge connecting heredity, environment and disease. In Parkinson's disease (PD), epigenetic regulation abnormalities have been widely reported. For example, DNA methylation is significantly altered in the enhancer region of neurons in PD patients and may be associated with dysregulation of alpha-synuclein (SNCA) gene expression\u003csup\u003e[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]\u003c/sup\u003e. Abnormalities in histone modifications, such as H3K9me2 and H3K27me3, have also been found to be associated with synaptic damage and neurodegeneration\u003csup\u003e[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]\u003c/sup\u003e. Epigenetic disorders accelerate the progression of PD by mediating the interaction between environmental and genetic factors. For example, SNCA gene hypomethylation increases α-synuclein expression and promotes pathological aggregation\u003csup\u003e[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]\u003c/sup\u003e. Mitochondrial genome epigenetic variation is associated with the speed of motor and cognitive decline in PD patients\u003csup\u003e[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]\u003c/sup\u003e. These studies reveal the multi-dimensional role of epigenetics in the whole process of PD disease.\u003c/p\u003e\u003cp\u003eIn this study, multiple omics data and machine learning models were integrated to identify PD-related epigenetic regulatory genes, and clinical samples were used to verify their expression and immune regulatory functions. The key genes and pathways identified provide a potential molecular basis for early diagnosis and targeted therapy of PD.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eData Sources and Preprocessing\u003c/h2\u003e\u003cp\u003eThe GSE7621 and GSE49036 datasets were downloaded from the Gene Expression Omnibus (GEO, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/geo/\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/geo/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e as training sets. Data normalization was performed using the Sangerbox tool, followed by batch effect removal via the Combat algorithm to generate combined datasets. The principal component analysis (PCA) cluster plot was used to visualize the effects of inter-batch difference removal.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eIdentification and Analysis of Differentially Expressed Genes (DEGs)\u003c/h3\u003e\n\u003cp\u003eTo analyze gene expression differences between healthy controls (HC) and Parkinson\u0026rsquo;s disease (PD) patients in substantia nigra tissues, the \"Limma\" (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://bioconductor.org/packages/release/bioc/html/limma.html\u003c/span\u003e\u003cspan address=\"https://bioconductor.org/packages/release/bioc/html/limma.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) package was applied to identify DEGs in the training sets using thresholds of |log2FoldChange| \u0026gt;1 and q-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05. The \"EnhancedVolcano\" R package was used to visualize DEGs in volcano plots. The top 50 significantly upregulated and downregulated genes in each sample were clustered based on log2FoldChange values, and heatmaps were generated using the \"pheatmap\" R package.\u003c/p\u003e\n\u003ch3\u003eConstruction of Weighted Gene Co-expression Network (WGCNA)\u003c/h3\u003e\n\u003cp\u003eTo identify PD-associated module genes, the \"WGCNA\" R package was used to construct a co-expression network for all samples in the training set. Hierarchical clustering analysis was performed to detect and remove outliers. A scale-free network was established, and genes were partitioned into modules. The correlation coefficient and p-value were calculated to analyze associations between modules and groups (PD vs. HC). Modules with the strongest correlations (|cor| \u0026gt;0.3 and p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) were selected, and genes within these modules were defined as PD-related module genes.\u003c/p\u003e\n\u003ch3\u003eIdentification of Candidate Genes\u003c/h3\u003e\n\u003cp\u003eTo identify PD-related genes associated with epigenetic factors, genes overlapping between DEGs, PD-related module genes, and epigenetic factor-related genes (from the EpiFactors database, v2.0) were extracted as candidate genes.\u003c/p\u003e\n\u003ch3\u003eFunctional Enrichment Analysis\u003c/h3\u003e\n\u003cp\u003eThe Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG)\u003csup\u003e[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]\u003c/sup\u003e and Reactome enrichment analyses of DEGs were conducted using the \u0026lsquo;clusterProfiler\u0026rsquo; (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://bioconductor.org/packages/release/bioc/html/clusterProfiler.html\u003c/span\u003e\u003cspan address=\"https://bioconductor.org/packages/release/bioc/html/clusterProfiler.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) package of R software. Findings of GO annotation analysis were categorized as follows: Biological Process (BP), Molecular Function (MF), and Cellular Components (CC)[1]. Significant enrichment was defined as adjusted P values\u0026thinsp;\u0026lt;\u0026thinsp;0.05.\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eProtein-Protein Interaction (PPI) Network Construction\u003c/h2\u003e\u003cp\u003eTo establish the interplay of DEGs, PPI network analysis was performed using the STRING database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cn.string-db.org/\u003c/span\u003e\u003cspan address=\"https://cn.string-db.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). A stringent interaction score threshold of 0.4 was employed to identify the most reliable and relevant interactions among the DEGs. Genes were denoted by nodes while connections between them were represented by edges. Subsequently, the main regulatory network was constructed and visualized using Cytoscape (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cytoscape.org/\u003c/span\u003e\u003cspan address=\"https://cytoscape.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The cytoHubba plugin of Cytoscape was used to identify the hub genes in the PPI network.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eMachine Learning\u003c/h3\u003e\n\u003cp\u003eTo systematically evaluate the predictive capacity of candidate genes for PD, 90 machine learning models were constructed based on 12 key genes, incorporating algorithms such as XGBoost, SVM, QDA, KNN, and Lasso regularization. Model performance was assessed using metrics including area under the curve (AUC), accuracy, recall, and F-score. Feature importance weights were calculated across models to generate gene importance bar plots and average rank plots, identifying genes with stable contributions. Visualizations (e.g., AUC curves, precision plots, and importance rankings) were used to illustrate model performance and gene relevance.\u003c/p\u003e\n\u003ch3\u003eValidation Set Expression Verification and Diagnostic Model Evaluation\u003c/h3\u003e\n\u003cp\u003eBox/violin plots were drawn to identify the differentially expressed hub genes between HC and PD patients in the verification set. Then, the efficacy of distinguishing between HC and PD patients was tested by receiver operating characteristic (ROC) curve analysis.\u003c/p\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eGene Set Enrichment Analysis (GSEA)\u003c/h2\u003e\u003cp\u003eGSEA was performed on the training set using the \"c2.cp.kegg.v7.0.symbols.gmt\" reference gene set. Spearman correlation coefficients between key genes and genome-wide genes were calculated using the \"psych\" R package. Enriched KEGG pathways were identified using the \"clusterProfiler\" R package (significance thresholds: p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.25, |NES| \u0026gt;1), and the top 10 significant pathways were visualized.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eImmune Cell Infiltration Analysis\u003c/h2\u003e\u003cp\u003eThe R package \u0026lsquo;CIBERSORT\u0026rsquo; (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/Moonerss/CIBERSORT\u003c/span\u003e\u003cspan address=\"https://github.com/Moonerss/CIBERSORT\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was applied to estimate immune cell infiltration abundance in PD and normal samples (excluding low-confidence samples with P\u0026thinsp;\u0026gt;\u0026thinsp;0.05). Associations between key gene expression (ACTL6B, NAP1L2, etc.) and immune cell infiltration were analyzed by dividing samples into high/low expression groups based on median values. Wilcoxon or t-tests (threshold: P\u0026thinsp;\u0026lt;\u0026thinsp;0.05) were used to compare immune cell abundance differences, visualized via boxplots.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003eReverse Transcription Quantitative Polymerase Chain Reaction (RT-qPCR)\u003c/h2\u003e\u003cp\u003eClinical whole blood samples were collected from both HC and PD patients, using heparinized EDTA anticoagulant tubes. The blood samples were centrifuged twice at 3000 g for 10 min each at 4\u0026deg;C. Serum was then harvested and stored at -80\u0026deg;C until further analysis.\u003c/p\u003e\u003cp\u003eTotal RNA was extracted using TRIzol reagent (Thermo Fisher Scientific, Waltham, Massachusetts, USA). RNA concentration was measured using a Nano-100 instrument (Allsheng, Hangzhou, China), and cDNA synthesis was performed using the Hiscript II QRT Supermix for qPCR kit (Vazyme, Nanjing, China). The reaction conditions were 50\u0026deg;C for 15 minutes and 85\u0026deg;C for 5 seconds. The qPCR experiments were executed on an ABI7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) with the following cycling parameters: pre-denaturation at 95\u0026deg;C for 30 seconds, denaturation at 95\u0026deg;C for 10 seconds, annealing at 60\u0026deg;C for 30 seconds, 40 cycles. GAPDH was used as an internal control. Ct values were analyzed using the 2\u003csup\u003e-ΔΔCt\u003c/sup\u003e method. The primer sequences were listed in Supplementary Table\u0026nbsp;1.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003eStatistical analysis\u003c/h2\u003e\u003cp\u003eBioinformatics statistical analyses were conducted using R language (version 4.0). All the data were processed by GraphPad Prism 7.0 statistical software (GraphPad, San Diego, CA, USA), and the data were presented as multiple groups of repeated data or means\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation. The comparisons between the two groups were analyzed by t-test. Statistical significance between two groups was determined by a t-test with a p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 considered significant.\u003c/p\u003e\u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003eIdentification of DEGs\u003c/h2\u003e\u003cp\u003eThe GSE7621 and GSE49036 datasets were utilized as training sets, comprising 16 PD patient substantia nigra tissue samples and 9 HC substantia nigra tissue samples, as well as 15 PD patient (Braak stages 3\u0026ndash;4 or 5\u0026ndash;6) substantia nigra tissue samples and 8 HC substantia nigra tissue samples, respectively. Standardization was performed using the Sangerbox tool, followed by batch effect removal via Combat to generate combined datasets, thereby minimizing potential biases from batch effects in subsequent analyses. The integrated dataset consisted of 31 PD patient substantia nigra samples and 16 HC substantia nigra samples. Visualization of density distributions (Supplementary Fig.\u0026nbsp;1A), UMAP distributions (Supplementary Fig.\u0026nbsp;1B), boxplots (Supplementary Fig.\u0026nbsp;1C), and principal component analysis (Supplementary Fig.\u0026nbsp;1D) demonstrated differences in data distribution before and after batch correction, confirming effective elimination of batch effects.\u003c/p\u003e\u003cp\u003eIn the training set samples, 78 DEGs between hc and PD patients were identified using thresholds of |log2FC| \u0026gt;1 and q-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05, and visualized in a volcano plot (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). Additionally, hierarchical clustering analysis was performed on the top 50 most significantly upregulated and downregulated genes in each sample, with the overall clustering results displayed in a heatmap (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\u003ch2\u003eIdentification of PD-Related Module Genes and Key Genes\u003c/h2\u003e\u003cp\u003eNext, WGCNA analysis and module identification were performed on the training set samples to identify PD-related module genes. First, hierarchical clustering analysis was conducted on all samples to detect outliers, and a clustering dendrogram was generated to reveal clear clustering patterns between the PD and HC groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). After removing outliers, the optimal soft threshold was determined using the pickSoftThreshold function to establish a scale-free network (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). Gene clustering and module identification were achieved via the \"dynamic tree cutting\" algorithm, generating distinct co-expression modules (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC). Module-trait relationship analysis demonstrated significant correlations between specific modules and disease status (PD vs. HC). Two key modules exhibited the strongest associations: the blue module (cor\u0026thinsp;=\u0026thinsp;0.37, p\u0026thinsp;=\u0026thinsp;0.01) showed a positive correlation with PD, while the turquoise module (cor\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.43, p\u0026thinsp;=\u0026thinsp;0.002) displayed a negative correlation (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD). Both modules met stringent thresholds (|cor| \u0026gt;0.3, p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), and genes within these modules were defined as PD-related module genes.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe PD-related module genes identified above were merged with DEGs obtained from prior Limma differential analysis, yielding 550 PD-associated genes. Subsequently, 796 non-redundant epigenetic factor-related genes were extracted from the EpiFactors database. Intersection analysis between these PD-associated genes and epigenetic factor genes revealed 12 overlapping genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE). These overlapping genes, representing PD-related epigenetic regulators, were defined as key genes for subsequent analyses.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\u003ch2\u003eFunctional Enrichment and PPI Network of PD-Related Genes\u003c/h2\u003e\u003cp\u003eTo systematically reveal the functional characteristics of PD-related genes, this study conducted multi-level functional annotation analysis on 550 PD-associated genes derived from the integration of DEGs and WGCNA. Functional enrichment analysis performed using the clusterProfiler package (p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05) demonstrated that the gene sets were significantly enriched in BP such as \"vesicle-mediated transport in synapse\" and \"regulated exocytosis\" (Supplementary Fig.\u0026nbsp;2A). CC analysis indicated that key genes were primarily localized to subcellular structures including \"synaptic membrane,\" \"transport vesicle,\" and \"transport vesicle membrane\" (Supplementary Fig.\u0026nbsp;2B). MF analysis revealed significant associations with biomolecular properties such as \"active monoatomic ion transmembrane transporter activity,\" \"calmodulin binding,\" and \"transmembrane transporter binding\" (Supplementary Fig.\u0026nbsp;2C). KEGG enrichment analysis further validated the significant enrichment of these genes in pathways such as \"Neurodegeneration-multiple diseases\" and \"Parkinson disease\" (Supplementary Fig.\u0026nbsp;2D). Reactome pathway analysis supplemented these findings by highlighting critical molecular mechanisms, including \"Transmission across Chemical Synapses\" and \"Neuronal System\" (Supplementary Fig.\u0026nbsp;2E). Notably, visualization of pathway interaction networks constructed via Cytoscape demonstrated that these enriched pathways formed distinct functional network modules through shared genes, with \"calcium-ion regulated exocytosis\" and \"inorganic cation transmembrane transporter activity\" acting as core hub nodes (Supplementary Fig.\u0026nbsp;3). This systematic multi-dimensional functional annotation provides novel molecular network insights into PD pathogenesis.\u003c/p\u003e\u003cp\u003eFurthermore, to explore protein-protein interactions encoded by PD-related genes, a PPI network was constructed using the STRING database and visualized with Cytoscape (Supplementary Fig.\u0026nbsp;4).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\u003ch2\u003eMachine Learning Evaluation of Key Genes\u003c/h2\u003e\u003cp\u003eThis study systematically evaluated the predictive efficacy of 12 key genes for PD by constructing a multidimensional evaluation framework based on 90 machine learning models. In terms of model performance, the AUC curves of the top 50 models (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA) revealed that the KNN algorithm (k\u0026thinsp;=\u0026thinsp;1/2) exhibited strong discriminatory power (AUC\u0026thinsp;\u0026gt;\u0026thinsp;0.7), while the XGBoost\u0026thinsp;+\u0026thinsp;Lasso combined model demonstrated stable performance in cross-gene prediction (average precision: 66%) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). Through Recall-Fscore dual-index analysis, QDA and multiple SVM variants (including SVM-default, SVM-CV:5, etc.) achieved optimal balance between model recall and prediction accuracy (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC-D).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eGene importance analysis highlighted significant algorithm-dependent characteristics. By integrating feature weights and average ranks across models, we identified 10 core predictive genes: EID2B, PRMT8, ACTL6B, SUDS3, SMYD3, EID2, YWHAZ, UBE2T, CDK5, and NAP1L2. Notably, EID2B was consistently recognized as a critical feature across all models, suggesting its stable predictive value transcending algorithms (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA-B). The integrated model performance metrics and gene importance analysis provide critical insights for further research.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\u003ch2\u003eValidation of Key Gene Expression\u003c/h2\u003e\u003cp\u003eTo systematically validate the reliability of key genes in the substantia nigra of PD patients, this study analyzed two independent cohorts from the GEO database. Wilcoxon rank-sum tests were performed to compare the expression of key genes in the validation sets: GSE20141 (10 PD patient SNpc neuron samples and 8 HC SNpc neuron samples) and GSE42966 (9 PD patient substantia nigra tissue samples and 6 HC samples) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA-B). Results showed discrepant differential expression patterns for PRMT8, SMYD3, and UBE2T between the two validation sets. Additionally, VAV1 expression data were absent in GSE42966.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eROC curve analysis evaluated the predictive performance of key genes. HSPA1A exhibited weak disease prediction capability (AUC\u0026thinsp;\u0026lt;\u0026thinsp;0.7) in both validation sets. Using an AUC threshold\u0026thinsp;\u0026gt;\u0026thinsp;0.7, three genes\u0026mdash;ACTL6B, NAP1L2, and PRMT8\u0026mdash;were identified as key epigenetics-related genes in the PD substantia nigra (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD), while BABAM1 and EID2 were highlighted as key genes associated with SNpc neurons and epigenetic regulation (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC).\u003c/p\u003e\u003cp\u003eTo further confirm bioinformatics findings, blood samples from PD patients (PD group) and healthy controls (HC group) were analyzed via RT-qPCR. The results showed that the mRNA expression levels of ACTL6B, NAP1L2, PRMT8, BABAM1, and EID2 were significantly decreased in the PD group (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e\u003ch2\u003eGSEA Enrichment Analysis\u003c/h2\u003e\u003cp\u003eTo elucidate the potential biological functions of key genes in PD pathogenesis, gene set enrichment analysis was performed using the KEGG database. Pathways meeting significance thresholds (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.25, |NES| \u0026gt;1) were prioritized. The top 10 enriched pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e) suggested that ACTL6B, NAP1L2, PRMT8, BABAM1, and EID2 may contribute to disease progression by regulating cellular metabolic reprogramming, proteasome activity, and related processes. These findings offer critical clues for deciphering molecular mechanisms.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec22\" class=\"Section2\"\u003e\u003ch2\u003eImmune Infiltration Analysis\u003c/h2\u003e\u003cp\u003eTo explore immune cell-PD interactions, CIBERSORT was applied to assess the infiltration abundance of 22 immune cell types in training set transcriptomic data (low-confidence samples with algorithm P\u0026thinsp;\u0026gt;\u0026thinsp;0.05 were excluded). Results revealed significant differences in CD8\u0026thinsp;+\u0026thinsp;T cells, mast cells, and macrophage subtypes (M1/M2) across samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eA), indicating immune microenvironment reprogramming in PD. Further analysis demonstrated significant correlations between key gene expression (ACTL6B, NAP1L2, PRMT8, BABAM1, EID2) and specific immune cell infiltration. After stratifying samples into high/low expression groups by median values, Wilcoxon rank-sum and independent t-tests showed that high ACTL6B expression correlated with elevated CD8\u0026thinsp;+\u0026thinsp;T cell and M2 macrophage infiltration, whereas PRMT8 and EID2 were linked to negative regulation of resting memory CD4\u0026thinsp;+\u0026thinsp;T cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eB-F).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eBiomarker research for PD is currently in a rapidly developing phase, providing assistance for critical clinical needs such as early diagnosis, progression monitoring, and therapeutic targets\u003csup\u003e[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]\u003c/sup\u003e. This study employed multi-dimensional integrated approaches including differential gene screening, Weighted Gene Co-expression Network Analysis (WGCNA), functional enrichment analysis, machine learning, Gene Set Enrichment Analysis (GSEA), and immune infiltration analysis, combined with external datasets, to screen key differentially expressed genes in PD. Partial gene expression was validated using RT-qPCR. This study found that epigenetic genes such as ACTL6B, PRMT8, NAP1L2, BABAM1 and EID2 were significantly related to the pathological process of PD. Key genes can regulate synaptic transmission, cellular metabolic reprogramming, and proteasome activity in multiple dimensions, providing new molecular network insights into the pathogenesis of PD. In addition, immunoinfiltration analysis showed changes in the abundance of immune cells in patients, indicating that the expression of key genes was significantly correlated with specific immune cell infiltration.\u003c/p\u003e\u003cp\u003eACTL6B belongs to the actin-associated protein family (ACTL6), which is involved in the chromatin remodeling complex\u003csup\u003e[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]\u003c/sup\u003e, regulates nucleosomal composition and gene transcription, and plays a key role in neuronal development and synaptic plasticity\u003csup\u003e[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]\u003c/sup\u003e. At present, the direct role of ACTL6B in Parkinson's disease has not been reported, but members of the chromatin remodeling complex (e.g. KAT8/KANSL1) regulate α-synuclein (SNCA) expression in PD through histone acetylation, thereby regulating parkinson-related genes\u003csup\u003e[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]\u003c/sup\u003e. It suggested that ACTL6B may indirectly affect neuronal survival by regulating neuronal differentiation and synaptic plasticity.\u003c/p\u003e\u003cp\u003eNAP1L2 is a nucleosome assembly protein, which is involved in the regulation of chromatin structure and gene expression as a histone chaperone\u003csup\u003e[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]\u003c/sup\u003e. Recent studies have found no direct link between NAP1L2 and PD, but found that its mRNA is regulated by m6A methylation, and m6A modifications (such as METTL3/ MeTTL14-mediated NAP1L2 methylation) are significantly dysregulated in PD, which may affect disease progression through regulation of neuronal gene expression\u003csup\u003e[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003ePRMT8, a member of the protein arginine methyltransferase family (PRMTs), is an important epigenetic regulatory enzyme\u003csup\u003e[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]\u003c/sup\u003e. PRMT8 plays a key role in shaping the morphology of dendritic spines by catalyzing the methylation of arginine residues and regulating the activity of target proteins\u003csup\u003e[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]\u003c/sup\u003e. Synaptic dysfunction is one of the early pathological features of Parkinson's disease, so PRMT8 may indirectly affect the survival and function of dopamine neurons by regulating synaptic structural stability.\u003c/p\u003e\u003cp\u003eBABAM1 is a member of the BRISC/BRCA1-A complex, which is involved in cell cycle regulation and inflammatory signaling pathways\u003csup\u003e[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]\u003c/sup\u003e. At present, the study does not mention the association or specific role of BABAM1 and PD. However, the BRISC complex is associated with neuroinflammation, which is an important pathological feature of PD\u003csup\u003e[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]\u003c/sup\u003e. Combined with the results of this study, it is speculated that it may affect PD progression through epigenetic-immune interaction network.\u003c/p\u003e\u003cp\u003eEID2 belongs to the EID gene family (E1A-like inhibitor of differentiation), which is usually involved in transcriptional regulation and cell differentiation\u003csup\u003e[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]\u003c/sup\u003e. At present, there are few direct studies on EID2. In focal injury of articular cartilage, EID2 expression is significantly higher than that in uninjured tissue, and its up-regulation may be related to inflammatory response, hypoxia or cellular stress\u003csup\u003e[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]\u003c/sup\u003e. Therefore, EID2 may be involved in the regulation of Parkinson's immune microenvironment by influencing the differentiation and function of immune cells.\u003c/p\u003e\u003cp\u003eThrough GO and KEGG enrichment analysis, it was found that PD-related genes were significantly enriched in \"vesicle-mediated transport in synapse\", \"calmodulin binding\" and \"transmembrane transporter\". GSEA enrichment analysis showed that ACTL6B, NAP1L2, PRMT8, BABAM1, EID2 and other key genes were involved in significant pathways including cell metabolic reprogramming and regulation of proteasome activity. According to the study of ManWK et al.\u003csup\u003e[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]\u003c/sup\u003e, the overexpression of α-Synuclein can interfere with the speed of vesicle refilling, destroy the circulation of synaptic vesicles, and lead to a decrease in the release of dopamine and other transmitters. Stavsky A et al.\u003csup\u003e[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]\u003c/sup\u003e found that the decreased mitochondrial calcium buffering capacity in PD led to the accumulation of reactive oxygen species and energy depletion, exacerbating abnormal synaptic transmission. These pathways are consistent with our findings. These suggest that PD-related genes play an important role in neurotransmitter release, synaptic transmission and cell metabolic.\u003c/p\u003e\u003cp\u003eIn this study, it was found that the infiltration abundance of immune cells such as CD8\u0026thinsp;+\u0026thinsp;T cells, mast cells and macrophage subtypes (M1/M2) in PD patients was significantly different. The expression of key genes is related to the infiltration of specific immune cells. In PD-related studies, the abnormality of immune cells and their association with genes have been reported many times. PD risk genes (such as LRRK2 and GBA) are highly expressed in monocytes and microglia and may affect disease risk by regulating inflammatory response\u003csup\u003e[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]\u003c/sup\u003e. The abnormal DNA methylation pattern of peripheral immune cells (such as CD14\u0026thinsp;+\u0026thinsp;monocytes and CD4\u0026thinsp;+\u0026thinsp;T cells) in PD patients may promote inflammation by regulating gene expression\u003csup\u003e[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]\u003c/sup\u003e. The key genes in this study were less involved in previous studies. This study found the correlation between key genes and immune cell infiltration, providing new clues for understanding the immunopathological mechanism of PD.\u003c/p\u003e\u003cp\u003eThis study is mainly based on bioinformatics analysis and clinical sample validation, and the specific molecular mechanisms of key genes regulating immune cell infiltration and pathway activity have not been clarified through cell or animal experiments. Although further mechanism verification is needed, this study lays a theoretical foundation for early diagnosis of PD and targeted epigenetic regulation to reshape the immune microenvironment. In the future, we will focus on the functional analysis of key genes, further elucidate their causal relationship in the occurrence and development of PD, and promote the transformation of basic research into clinical applications.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study adopted integrated multi-omics and machine learning methods, combined with differential gene screening, WGCNA, GSEA and immune infiltration analysis, and verified by RT-qPCR, successfully identified epigenetic and immune hub genes associated with PD, revealing their links with immune microenvironment and key pathways, and laying a theoretical foundation for early diagnosis and precise treatment of PD.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cdiv id=\"Sec25\" class=\"Section2\"\u003e\u003ch2\u003eCorrespondence\u003c/h2\u003e\u003cp\u003eGuiyue Meng, the Second Hospital of Shandong University, Jinan, Shandong, China. E-mail: [email protected].\u003c/p\u003e\u003c/div\u003e\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study was approved by the Ethics Committee of The Second Hospital of Shandong University. All principles in the Declaration of Helsinki were strictly followed in this study. The specimens used for RT-qPCR validation were obtained from inpatients. All patients provided signed informed consent forms, authorizing further research on blood specimens. All cited statistics from published studies and publicly available database were approved by a relevant review board.\u003c/p\u003e\n\u003ch2\u003eClinical trial registration\u003c/h2\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003eThis study did not involve clinical trials or patient interventions, as it primarily analyzed publicly available transcriptomic datasets and institutional biobank samples with informed consent.\u003c/p\u003e\n\u003ch2\u003eConsent for publication\u003c/h2\u003e\n\u003cp\u003eThe data used in this article are from publicly available databases, and the source databases have obtained informed consent to publish the researches and any accompanying images without disclosing personal information. No personal information was disclosed in this study.\u003c/p\u003e\n\u003ch2\u003eCompeting interests\u003c/h2\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003ch2\u003eAuthors information\u003c/h2\u003e\n\u003cp\u003eAuthor List: Guiyue Meng\u003csup\u003e* 1,\u003c/sup\u003e, Yunqi Sun \u003csup\u003e1,\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eAffiliate Institutions:\u003c/p\u003e\n\u003ch2\u003eFunding\u003c/h2\u003e\n\u003cp\u003eNone.\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eGuiyue Meng analyzed the data and wrote the manuscript. Yunqi Sun edited the manuscript and made suggestions for revision. All authors declare no disputes of interest.\u003c/p\u003e\n\u003ch2\u003eAcknowledgements\u003c/h2\u003e\n\u003cp\u003eOur sincere thanks to all participants included in this study and developers of the R software and all the related algorithms and websites.\u003c/p\u003e\n\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eThe GSE7621 and GSE49036 datasets were downloaded from the Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/) as training sets.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eDurmaz Celik N, Ozben S, Ozben T. Unveiling Parkinson's disease through biomarker research: current insights and future prospects. Crit Rev Clin Lab Sci. 2024;61(7):529\u0026ndash;45. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/10408363.2024.2331471\u003c/span\u003e\u003cspan address=\"10.1080/10408363.2024.2331471\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang D, Zhang J, Wang Y, et al. Targeting epigenetic modifications in Parkinson's disease therapy. Med Res Rev. 2023;43(5):1748\u0026ndash;77. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/med.21962\u003c/span\u003e\u003cspan address=\"10.1002/med.21962\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMourtzi T, Kazanis I. Endogenous\u0026thinsp;\u0026lt;\u0026thinsp;i\u0026thinsp;\u0026gt;\u0026thinsp;versus\u0026thinsp;exogenous cell replacement for Parkinson's disease: where are we at and where are we going? Neural Regen Res. 2022;17(12):2637\u0026ndash;42. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.4103/1673-5374.336137\u003c/span\u003e\u003cspan address=\"10.4103/1673-5374.336137\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCao Y, Xu Y, Cao M, et al. Fluid-based biomarkers for neurodegenerative diseases. Ageing Res Rev Published online March. 2025;21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.arr.2025.102739\u003c/span\u003e\u003cspan address=\"10.1016/j.arr.2025.102739\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang D, Zhang J, Wang Y, et al. Targeting epigenetic modifications in Parkinson's disease therapy. Med Res Rev. 2023;43(5):1748\u0026ndash;77. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/med.21962\u003c/span\u003e\u003cspan address=\"10.1002/med.21962\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGuhathakurta S, Kim J, Adams L, et al. Targeted attenuation of elevated histone marks at SNCA alleviates α-synuclein in Parkinson's disease. EMBO Mol Med. 2021;13(2):e12188. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.15252/emmm.202012188\u003c/span\u003e\u003cspan address=\"10.15252/emmm.202012188\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang Z, Wang R, Zhou H et al. Inhibition of EHMT1/2 rescues synaptic damage and motor impairment in a PD mouse model. Cell Mol Life Sci. 2024;81(1):128. Published 2024 Mar 12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00018-024-05176-5\u003c/span\u003e\u003cspan address=\"10.1007/s00018-024-05176-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ede Mena L, Par\u0026eacute;s G, Garrido A, et al. α-Synuclein Gene Hypomethylation in LRRK2 Parkinson's Disease Patients. Mov Disord. 2025;40(3):550\u0026ndash;5. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/mds.30094\u003c/span\u003e\u003cspan address=\"10.1002/mds.30094\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu G, Ni C, Zhan J, et al. Mitochondrial haplogroups and cognitive progression in Parkinson's disease. Brain. 2023;146(1):42\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/brain/awac327\u003c/span\u003e\u003cspan address=\"10.1093/brain/awac327\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKanehisa M, Furumichi M, Sato Y, Matsuura Y, Ishiguro-Watanabe M. KEGG: biological systems database as a model of the real world. Nucleic Acids Res. 2025;53(D1):D672\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/nar/gkae909\u003c/span\u003e\u003cspan address=\"10.1093/nar/gkae909\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang L, Liu Y, Lu Y, Wang G. Targeting epigenetics as a promising therapeutic strategy for treatment of neurodegenerative diseases. Biochem Pharmacol. 2022;206:115295. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.bcp.2022.115295\u003c/span\u003e\u003cspan address=\"10.1016/j.bcp.2022.115295\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBell S, Rousseau J, Peng H, et al. Mutations in ACTL6B Cause Neurodevelopmental Deficits and Epilepsy and Lead to Loss of Dendrites in Human Neurons. Am J Hum Genet. 2019;104(5):815\u0026ndash;34. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ajhg.2019.03.022\u003c/span\u003e\u003cspan address=\"10.1016/j.ajhg.2019.03.022\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCali E, Quirin T, Rocca C, et al. Clinical and genetic delineation of autosomal recessive and dominant ACTL6B-related developmental brain disorders. Genet Med Published online September. 2024;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.gim.2024.101251\u003c/span\u003e\u003cspan address=\"10.1016/j.gim.2024.101251\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHicks AR, Reynolds RH, O'Callaghan B, et al. The non-specific lethal complex regulates genes and pathways genetically linked to Parkinson's disease. Brain. 2023;146(12):4974\u0026ndash;87. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/brain/awad246\u003c/span\u003e\u003cspan address=\"10.1093/brain/awad246\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHu M, Xing L, Zhang L, et al. NAP1L2 drives mesenchymal stem cell senescence and suppresses osteogenic differentiation. Aging Cell. 2022;21(2):e13551. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/acel.13551\u003c/span\u003e\u003cspan address=\"10.1111/acel.13551\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYan J, Wang Z, Li Y, Li R, Xiang K. m6A-related genes and their role in Parkinson's disease: Insights from machine learning and consensus clustering. Med (Baltim). 2024;103(45):e40484. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/MD.0000000000040484\u003c/span\u003e\u003cspan address=\"10.1097/MD.0000000000040484\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHsu SH, Hung WC. Protein arginine methyltransferase 3: A crucial regulator in metabolic reprogramming and gene expression in cancers. Cancer Lett. 2023;554:216008. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.canlet.2022.216008\u003c/span\u003e\u003cspan address=\"10.1016/j.canlet.2022.216008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLo LH, Dong R, Lyu Q, Lai KO. The Protein Arginine Methyltransferase PRMT8 and Substrate G3BP1 Control Rac1-PAK1 Signaling and Actin Cytoskeleton for Dendritic Spine Maturation. Cell Rep. 2020;31(10):107744. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.celrep.2020.107744\u003c/span\u003e\u003cspan address=\"10.1016/j.celrep.2020.107744\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJin F, Zhu Y, Liu M, et al. Babam2 negatively regulates osteoclastogenesis by interacting with Hey1 to inhibit Nfatc1 transcription. Int J Biol Sci. 2022;18(11):4482\u0026ndash;96. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7150/ijbs.72487\u003c/span\u003e\u003cspan address=\"10.7150/ijbs.72487\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Published 2022 Jul 11.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang W, Liu K, Ren GM, et al. BRISC is required for optimal activation of NF-κB in Kupffer cells induced by LPS and contributes to acute liver injury. Cell Death Dis. 2023;14(11):743. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41419-023-06268-z\u003c/span\u003e\u003cspan address=\"10.1038/s41419-023-06268-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Published 2023 Nov 15.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFu X, Luo L, Yi R, et al. Transcriptome profiling in Eid1-KO mice brain shows that Eid1 links cell proliferation in the brain. Gene. 2019;717:143998. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.gene.2019.143998\u003c/span\u003e\u003cspan address=\"10.1016/j.gene.2019.143998\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAşık MD, G\u0026uuml;rsoy S, Akkaya M, Kozacı LD, Doğan M, Bozkurt M. Microarray analysis of cartilage: comparison between damaged and non-weight-bearing healthy cartilage. Connect Tissue Res. 2020;61(5):456\u0026ndash;64. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/03008207.2019.1611797\u003c/span\u003e\u003cspan address=\"10.1080/03008207.2019.1611797\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMan WK, Tahirbegi B, Vrettas MD, et al. The docking of synaptic vesicles on the presynaptic membrane induced by α-synuclein is modulated by lipid composition. Nat Commun. 2021;12(1):927. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41467-021-21027-4\u003c/span\u003e\u003cspan address=\"10.1038/s41467-021-21027-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Published 2021 Feb 10.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eStavsky A, Stoler O, Kostic M et al. Aberrant activity of mitochondrial NCLX is linked to impaired synaptic transmission and is associated with mental retardation [published correction appears in Commun Biol. 2021;4(1):755. doi: 10.1038/s42003-021-02312-w.]. Commun Biol. 2021;4(1):666. Published 2021 Jun 2. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s42003-021-02114-0\u003c/span\u003e\u003cspan address=\"10.1038/s42003-021-02114-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNavarro E, Udine E, Lopes KP, et al. Dysregulation of mitochondrial and proteolysosomal genes in Parkinson's disease myeloid cells. Nat Aging. 2021;1(9):850\u0026ndash;63. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s43587-021-00110-x\u003c/span\u003e\u003cspan address=\"10.1038/s43587-021-00110-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAndersen MS, Leikfoss IS, Brorson IS et al. Epigenome-wide association study of peripheral immune cell populations in Parkinson's disease. NPJ Parkinsons Dis. 2023;9(1):149. Published 2023 Oct 31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41531-023-00594-x\u003c/span\u003e\u003cspan address=\"10.1038/s41531-023-00594-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Parkinson’s disease, differentially expressed genes, machine learning, immune infiltration, epigenetic regulation","lastPublishedDoi":"10.21203/rs.3.rs-7376065/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7376065/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eParkinson's disease (PD) is a neurodegenerative disorder, and its early diagnosis and treatment remain significant challenges. Identifying molecular biomarkers associated with PD pathogenesis is critical for disease intervention.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eThe GSE7621 and GSE49036 datasets were integrated to identify differentially expressed genes (DEGs) between PD patients' substantia nigra tissues and normal samples. Weighted Gene Co-expression Network Analysis (WGCNA) was employed to screen PD-related module genes. DEGs were cross-analyzed with epigenetic factor-related genes to identify key genes. The predictive efficacy of these key genes was evaluated using 90 machine learning models, and their expression was validated via independent GEO datasets (GSE20141, GSE42966) and reverse transcription-quantitative polymerase chain reaction (RT-qPCR). CIBERSORT was utilized to analyze immune cell infiltration, and Gene Set Enrichment Analysis (GSEA) was performed to explore the biological functions and molecular mechanisms of key genes.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eA total of 78 DEGs were identified, and 550 PD-related genes were screened through WGCNA, yielding 12 key genes. Machine learning models revealed ACTL6B, PRMT8, NAP1L2, BABAM1, and EID2 as core predictive genes for PD. Independent validation and RT-qPCR confirmed significant downregulation of ACTL6B, PRMT8, NAP1L2, BABAM1, and EID2 in PD patients. Immune infiltration analysis demonstrated altered infiltration abundance of CD8\u0026thinsp;+\u0026thinsp;T cells, macrophage subtypes, and other immune cells in PD patients, with key genes linked to immune microenvironment regulation. GSEA indicated that these genes participate in pathways such as cellular metabolic reprogramming and synaptic transmission.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e\u003cp\u003eThis study systematically identified PD-associated epigenetic regulatory genes and revealed their connections to immune microenvironment dynamics and molecular pathways, offering novel insights for early PD diagnosis and therapeutic target development.\u003c/p\u003e","manuscriptTitle":"Correlation analysis and clinical validation of Parkinson's disease and epigenetic factor-related genes based on transcriptome data","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-25 01:32:16","doi":"10.21203/rs.3.rs-7376065/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"cbfdfb02-c8cd-480d-a0b8-fa07762ef4eb","owner":[],"postedDate":"September 25th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-10-09T07:40:06+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-25 01:32:16","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7376065","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7376065","identity":"rs-7376065","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}