{"paper_id":"0603961a-e804-4e95-b775-de2b21bd0564","body_text":"Identification of key genes and regulatory networks associated with atherosclerotic carotid artery stenosis through comprehensive bioinformatics analysis and machine learning | 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 Identification of key genes and regulatory networks associated with atherosclerotic carotid artery stenosis through comprehensive bioinformatics analysis and machine learning Lei Yang, Peidong He, Wenkang Lei, Zhen Wang, Jingjing Liang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7112702/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 04 Nov, 2025 Read the published version in European Journal of Medical Research → Version 1 posted 11 You are reading this latest preprint version Abstract Objective: A comprehensive bioinformatics analysis was conducted to identify key genes and regulatory networks associated with atherosclerotic carotid artery stenosis (ACAS). Methods: Four datasets, including GSE43292, GSE100927, GSE28829, and GSE198600, were integrated to form the training set, with the GSE163154 dataset serving as the validation set. Subsequently, differential expression and functional enrichment analysis were performed on the training set. Additionally, key pathogenic genes were identified using the protein-protein interaction networks, molecular complex detection technique, and three machine learning (ML) algorithms. These identified genes were validated through inter-group differences and receiver operating characteristic (ROC) curve analyses. Immune-related functions and immune cell correlations were analyzed and verified using ACAS plaque tissue samples. Results: Following the analysis, a total of 33 downregulated and 52 upregulated genes were identified. Furthermore, enrichment analysis of gene sets demonstrated that the highly expressed group was involved in cellular receptor signaling, leishmaniasis infection, lysosome, PPAR-signaling, and Toll-like receptor pathways. In contrast, the low-expressed group was involved in mechanisms involving dilated cardiomyopathy, pyruvate metabolism, hypertrophic cardiomyopathy, spliceosome, and TGF-β signaling pathways. Notably, ANPEP, CSF1R, MMP9, and CASQ2 were found to differ significantly between groups. Correlation analysis revealed positive associations between MMP9 expression and neutrophil infiltration, CASQ2 expression and M2 macrophage abundance, and CSF1R expression and M1 macrophage levels. Conclusion: Consequently, these genes may serve as potential biomarkers and therapeutic targets in the diagnosis and treatment of ACAS. Atherosclerotic carotid artery stenosis Pathogenic markers Immune cell correlation Machine learning Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 1. Introduction Atherosclerotic carotid artery stenosis (ACAS), which is the narrowing of the carotid artery due to the formation of atherosclerotic plaque, is a highly prevalent condition. This condition may progress from partial stenosis to complete occlusion, leading to adverse events such as ischemic stroke (IS) or transient ischemic attack[ 1 ]. The prevalence of ACAS is increasing due to population aging and increased incidences of cerebrovascular risk factors. Associated stroke events have imposed a significant economic burden on both the families and the broader healthcare system[ 2 ]. Consequently, there is an urgent need to investigate its causative factors, to facilitate the implementation of preventive and control measures to improve the prognosis of ACAS patients. Patients with ACAS are at increased risk of developing stroke. Typically, most of them require carotid endarterectomy (CEA) or stenting to improve cerebral blood flow, however, these therapies are not effective in completely treating ACAS, with many patients subsequently developing restenosis[ 3 ]. Notably, CEA is associated with approximately 2–20% incidence of postoperative stroke—a rare and severe complication[ 4 ]. While existing pharmacological agents are primarily used to manage blood pressure and atherothrombosis, the genetic and molecular mechanisms underlying ACAS remain incompletely understood[ 5 ]. Consequently, in-depth exploration of these molecular mechanisms offer a promising potential for the development of novel therapeutic strategies for ACAS. Notably, ACAS is a regulatory condition that arises from key inflammatory, immune, and metabolic pathways. Research has shown that reversing the pathogenesis of ACAS, along with improving prognostic outcomes, has been a challenging task[ 6 ]. However, machine learning (ML) offers a promising avenue towards this goal. Notably, ML is characterized by reduced error rates, with superior predictive performance, thereby enhancing its widespread adoption for target screening, prognosis prediction, and providing guidance for optimal therapeutic option[ 7 , 8 ]. This study used an integrated approach involving protein-protein interaction (PPI) network analysis and molecular complex detection (MCODE) to explore the molecular relationships among key genes involved in ACAS. Machine learning models, including least absolute shrinkage and selection operator (LASSO), random forest (RF), and support vector machine recursive feature elimination (SVM-RFE), were utilized for a comprehensive analysis to identify critical targets. These identified targets may provide a valuable therapeutic foundation for developing precise risk prediction models and personalized treatment decisions in ACAS. 2. Materials and methods 2.1 Retrieval and merging of datasets The ACAS datasets used in this study were retrieved from the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) database and they included the following: GSE100927 (12 normal endothelium samples + 29 carotid atherosclerosis samples), GSE28829 (13 early plaque samples + 16 advanced carotid atherosclerosis samples), GSE198600 (5 stable carotid atherosclerotic plaque samples + 6 unstable carotid atherosclerosis samples), GSE43292 (32 normal endothelium samples + 32 carotid atherosclerosis samples), and GSE163154 (16 stable carotid atherosclerosis samples + 27 unstable carotid atherosclerosis samples). Notably, the first four datasets were merged to form the training set (n = 145), with GSE163154 (n = 43) serving as the validation set. Details of the datasets are listed in Table 1 . Batch correction was performed on the datasets using the “sva” package (version 3.50.0), while the principal component analysis (PCA) was used to visualize the results before and after correction. Table 1 Basic information on the GEO datasets used in the study. ID GSE series Disease Samples Platform Group 1 GSE43292 CAS 32 carotid atherosclerosis plaques and 32 endosomal tissues GPL6244 Discovery set 2 GSE100927 CAS 29 carotid atherosclerosis plaques and 12 endosomal tissues GPL17077 Discovery set 3 GSE198600 CAS 6 unstable carotid plaque samples and 5 stable carotid plaque samples GPL11154 Discovery set 4 GSE28829 CAS 16 advanced carotid plaque samples and 13 early carotid plaque samples GPL570 Discovery set 5 GSE163154 CAS 27 unstable plaque samples and 16 stable plaque samples GPL6104 validation set 2.2 Clinical samples of ACAS A total of 20 ACAS patients admitted at Renmin Hospital of Wuhan University and 10 ACAS patients admitted at The First Affiliated Hospital of Yangtze University between 2021 and 2023 were enrolled in this study. Both the control and experimental groups included 15 ACAS patients who underwent CEA. The experimental group comprised the atherosclerotic plaque tissues, whereas the control group comprised the adjacent non-plaque intimal tissues. The atherosclerotic plaque and adjacent non-plaque intimal tissues were collected after the CEA. The study protocol was approved by the Clinical Research Ethics Committee of Renmin Hospital of Wuhan University (Approval No. WDRY2023-K123) and the ethics committee of the First Affiliated Hospital of Yangtze University (Approval No. LL202301). Additionally, the study methodologies strictly adhere to the relevant guidelines and regulations. All participants signed the informed consent form. 2.3 Identification of differentially expressed genes (DEGs) The “limma” package (version 3.60.4) was used to identify the DEGs by performing normalization and analysis of gene expression patterns. The criteria for identifying the DEGs were based on an adjusted p-value < 0.05 and an absolute log 2 fold change (|logFC|) > 1. The results were visualized by generating heatmaps using the “pheatmap” R package (version 1.0.12). 2.4 Functional enrichment analysis Functional and pathway enrichment analyses of the DEGs were performed with the “ClusterProfiler,” “enrichplot,” and “org.Hs.eg.db” packages. Gene ontology (GO)[ 9 ] and Kyoto encyclopedia of genes and genomes (KEGG)[ 10 ] were employed for detailed annotation and pathway enrichment analyses. The reference genome file “c2.cp.kegg.Hs.symbols.gmt” was used to conduct the gene set enrichment analysis (GSEA)[ 11 ] to distinguish the differences in the associated pathways. All results were visualized using the “ggplot2” R package (version 3.5.1). 2.5 Protein-protein interaction (PPI) and MCODE analysis The PPI network of the 85 DEGs was constructed by uploading them into the STRING database ( http://string-db.org ) with a medium confidence level of 0.400. The PPI network was visualized using the Cytoscape_3.10.0 (version 3.10.0) software. To identify key sub-networks and core genes, MCODE[ 12 ], analysis was conducted within Cytoscape, facilitating the extraction of critical sub-networks and core genes based on the topological relationships between edges and nodes within the PPI network. Furthermore, functional module extraction and node information analysis were performed using MCODE with specific parameter settings as follows: Degree Cutoff of 2, Node Score Cutoff of 0.2, K-Core of 2, and a Max Depth from Seed of 100. 2.6 Three Machine learning methods for filtering feature genes Pathogenic genes were identified from the functional modules using three approaches: LASSO, SVM-RFE, and Random Forest. Notably, the LASSO regression was used to generate coefficient paths and cross-validation curves to determine optimal λ values and characteristic gene numbers. The SVM-RFE evaluation plots illustrated the performance of feature selection based on 10-fold cross-validation accuracy and error rates. The RF algorithm identified optimal tree numbers (500 trees) via error minimization, with top-ranked genes selected based on importance scores for further analysis. The overlapping results from the three algorithms were identified, and a Venn diagram was constructed using the “Venn Diagram” R package (version 1.7.3). Furthermore, ROC curves were plotted, and then the area under the curve (AUC) was calculated with the “pROC” and “InpROC” packages in R software (version 1.18.5) to determine the predictive value of feature genes in both the training and validation sets. 2.7 Construction of carotid stenosis maps A nomogram was constructed based on the identified feature genes using the “rms” (version 6.8) and “rmda” (version 1.6) packages in R. Visual assessment of the predictive performance of the nomogram was evaluated by generating calibration curves. Clinical impact of the nomogram was further assessed by generating clinical impact curves, which illustrate the potential clinical applicability of the model. Furthermore, the clinical utility and net benefit of the nomogram was assed using the decision curve analysis (DCA), thereby ensuring its clinical relevance and value in practical decision-making scenarios. 2.8 Immune-related function and immune cell correlation analysis A total of 83 carotid atherosclerosis samples were stratified into high-expression groups based on the expression levels of the target genes. The differences in immune-related functions between the two groups were performed through functional differential analysis using the R packages “GSVA” (version 1.52.3) and “GSEABase” (version 1.66.0). Furthermore, correlation analysis was conducted to investigate the associations between feature genes and immune cells, using the “ggpubr” (version 0.6.0), “reshape2” (version 1.4.4), and ‘ggExtra” (version 0.10.1) packages for visualization and comprehensive analysis. 2.9 Reverse transcription quantitative polymerase chain reaction (RT-qPCR) The tissue samples were lysed using the Trizol reagent. Subsequently, RevertAidTM First Strand cDNA Synthesis Kit (Takara, Japan) was used to generate complementary DNA with the following reaction conditions: exposure to room temperature for 10 mins, 42°C for 60 mins, 95°C for 5 mins, and freezing for 5 mins. Using SYBR Green reagent, the RT-qPCR reaction was conducted using the 7500 PCR equipment. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal reference to normalize target gene expression levels, which were quantified using the 2 −ΔΔct technique. The primer sequences used are shown in Table S1 . 2.10 Immunofluorescence (IF) staining The tissue samples were fixed in 10% formalin, serially dehydrated in ethanol at increasing percentages, and then incubated in chloroform. Subsequently, the fixed tissues were embedded in paraffin for double immunofluorescent labeling. Following antigen retrieval using hot citrate buffer, the deparaffinized sections were then soaked in a blocking reagent for 1 hour. The primary antibodies were incubated with human carotid atherosclerotic plaques at 4°C overnight. The primary antibodies used included: CD68 (servicebio, GB113150), CD206 (servicebio, GB115273), CASQ2 (servicebio, GB114847), CSF1R (servicebio, GB11581), MPO (servicebio, GB11224), and MMP9 (servicebio, GB12132). The tissue sections were thoroughly washed in readiness for secondary antibody incubation, which involved 2 hours of incubation with Alexa Fluor secondary antibodies from Thermo Fisher Scientific. The stained tissue sections were then mounted using Vectashield Mounting Medium (Burlingame, CA, USA) to preserve fluorescence and ensure high-quality visualization. Furthermore, tissue sections were treated with 2 mol/L HCl for 20 mins at 30°C to denature DNA for BrdU staining. Subsequently, the sections were rinsed with PBS and incubated with 0.1 mol/L borate buffer for 10 mins at room temperature to neutralize the acid. Following an additional PBS wash, the slices were subjected to the previously described immunofluorescence staining procedure. 2.11 Construction of regulatory network involving characteristic genes, competing endogenous RNA (ceRNA), and transcription factors (TFs) To identify miRNAs potentially interacting with the feature genes, miRNA prediction was performed using miRDB, miRanda, and TargetScan techniques. Subsequent analysis involved only miRNAs commonly identified by all three tools. Long non-coding RNA (lncRNA) interacting with the identified miRNAs were analyzed using the SpongeScan database to identify potential miRNA-binding lncRNAs[ 13 ]. Subsequently, a ceRNA regulatory network was constructed based on these interaction results and visualized using the Cytoscape software, integrating the interactions between lncRNAs, miRNAs, and feature genes. Additionally, a gene-TF regulatory network was constructed using the NetworkAnalyst ( http://www.networkanalyst.ca )[ 14 ], thereby providing a comprehensive analysis visualization of the regulatory mechanisms associated with the feature genes. 2.12 Statistical analysis All statistical analyses were performed using the R software (version 4.4.0). Normally distributed variables were analyzed using the t-test, while the Wilcoxon test was used for non-normally distributed variables. Correlation analyses were conducted using the Pearson analysis for linear relationships, while the Spearman analysis was used for monotonic relationships. All statistical tests were two-sided, with an adjusted p -value < 0.05 considered to be statistically significant. 3. Results 3.1 A total of 85 DEGs were identified Batch correction was performed before any differential analysis. Notably, the PCA analysis revealed that the samples from different datasets were distinct before the correction, indicating that there was a batch effect between these samples (Fig. 1 A). However, after batch correction, these samples were distributed randomly, indicating elimination of the batch effect (Fig. 1 B). All gene volcanoes were then plotted (Fig. 1 C), and differential analysis identified 85 DEGs, with 33 downregulated and 52 upregulated genes (Fig. 1 D). 3.2 Functional and pathway analysis of the 85 DEGs Functional enrichment analysis was conducted on the 85 DEGs. Notably, the KEGG and GO analysis revealed that these genes were mainly associated with negative regulation of transport activities, myeloid leukocyte differentiation, and leukocyte chemotaxis (Fig. 2 A), as well as pathways such as phagolysosomes, tuberculosis, diabetic cardiomyopathy, lipids and atherosclerosis (Fig. 2 B). Furthermore, GSEA analysis was conducted to identify the differences in the pathways associated with the high and low-expressed genes. The results revealed that dilated cardiomyopathy, pyruvate metabolism, hypertrophic cardiomyopathy, spliceosome, and TGF-β signaling pathways were highly activated in the low-expressed genes (Fig. 2 C), whereas cellular receptor signaling pathway, leishmaniasis infection, lysosome, PPAR-signaling pathway, and Toll -like receptor signaling pathway were significantly activated in the high-expressed genes (Fig. 2 D). 3.3 Three important functional modules contained 26 genes using MCODE analysis The PPI map of the 85 DEGS was constructed (Fig. 3 A). To investigate the potential molecular mechanisms underlying ACAS, a modular network analysis was performed using the MCODE algorithm to identify densely connected regions within the PPI network. This analysis revealed three key modules comprising a total of 26 genes, representing potential core therapeutic targets for ACAS (Fig. 3 B-D). The detailed information of each module is shown in Table 2 . Table 2 The results of the MCODE analysis. Cluster Nodes Edges Node IDs 1 18 101 TREM1, ITGAM, MMP9, CCR1, LY86, C1QB, CTSS, CD52, ITGB2, SPP1, NCF2, CD36, APOE, CD14, CSF1R, LAPTM5, ANPEP, TYROBP 2 5 9 MYOCD, ATP1A2, PLN, RYR2, CASQ2 3 3 3 MMP12, MMP1, CHI3L1 3.4 Four feature genes were identified using the three ML algorithms The LASSO algorithm identified a total of 9 feature genes—including MMP9, CCR1, CD14, CSF1R, ANPEP, MYOCD, ATP1A2, CASQ2, and CHI3L1 (Fig. 4 A, B). The SVM-RFE analysis on the 26 genes identified 11 feature genes—including PLN, ANPEP, CSF1R, CTSS, MYOCD, C1QB, APOE, CHI3L1, MMP9, ITGAM, and CASQ2 (Fig. 4 C, D). Additionally, the RF method identified the top 10 feature genes based on relative importance, including ITGB2, APOE, MMP1, SPP1, MMP9, CASQ2, CSF1R, C1QB, NCF2, and ANPEP (Fig. 4 E, F). Subsequently, the intersection of the identified genes was identified as four common feature genes: ANPEP, CASQ2, CSF1R, and MMP9 (Fig. 4 G). 3.5 Intergroup differences among the four feature genes To visualize the differences in the expression levels among the four identified genes, line and violin plots were generated. The results revealed significant differences between the expression of the four feature genes in both the experimental and control groups (p < 0.01, Fig. 5 A, B). Additionally, ANPEP, CSF1R, and MMP9 were upregulated, while CASQ2 was downregulated in the experimental group (Fig. 5 C). 3.6 ROC analysis of the 4 signature genes To verify the diagnostic accuracy of the identified genes, ROC analysis was conducted. In the training set, the AUC values for ANPEP, CASQ2, CSF1R, and MMP9 were 0.851, 0.864, 0.914, and 0.861, respectively (Fig. 6 A). In the validation set, the AUC values for ANPEP, CASQ2, CSF1R, and MMP9 were 0.947, 0.921, 0.944 and 0.935, respectively (Fig. 6 B). 3.7 Association between the prevalence of ACAS and characteristic genes A column chart was constructed to illustrate the diagnostic performance of the combined expression levels of the four signature genes in identifying ACAS (Fig. 7 A). The calibration curve demonstrated high accuracy in predicting the disease prevalence ( Fig. 7 B ) , while the clinical impact curve validate that the model had a robust predictive capability across various thresholds ( Fig. 7 C ) . Furthermore, DCA indicated significant clinical benefits for patients diagnosed with ACAS using the nomogram. This indicated that the nomogram could be utilized as a reliable diagnostic tool in clinical practice ( Fig. 7 D ) . 3.8 Immune-related function and immune cell correlation analysis of the characteristic genes There were significant differences in immune functionality between the high- and low-expression groups as evidenced by the immune-related function and immune cell correlation analysis of the identified genes ( Fig. 8 A-D ) . This observation indicates that the identified genes significantly impact the immune microenvironment. Notably, ANPEP was positively correlated with immune-suppressive cells, such as Tregs and resting NK cells, while negatively correlated with antigen-presenting and cytotoxic cells, such as CD8 + T and dendritic cells ( Fig. 8 E ) . Additionally, CASQ2 influenced macrophage polarization, revealing high propensity towards the anti-inflammatory M2 macrophages over pro-inflammatory M0 macrophages ( Fig. 8 F ) . Also, CSF1R modulated macrophage activity, positively correlating with M0 macrophages and inversely with activated NK cells and M2 macrophages ( Fig. 8 G ) . The MMP9 exhibited associations with neutrophil and resting NK cell activity, while inversely correlated with activated CD8 + T cells and NK cells ( Fig. 8 H ) , suggesting a role in immune surveillance and inflammation. Collectively, these findings highlight the complex interactions between gene expression patterns and immune cell dynamics, providing valuable insights into their contributions to the modulation of the immune microenvironment. 3.9 Validation of the identified genes using tissue samples The expression levels of the identified genes were assessed between the carotid atherosclerotic plaque tissues and adjacent non-plaque intimal tissues. Notably, RT-qPCR analysis revealed that the levels of MMP9, ANPEP, and CSF1R were upregulated, while CASQ2 was downregulated (Fig. 9 A). The immune cell correlations among the identified genes in ACAS were evaluated using the IF double-labeling technique on carotid atherosclerotic plaques from patients undergoing CEA. The following results were observed: positive correlations between CASQ2 and M2 macrophages (CD206), CSF1R and M1 macrophages (CD86), and MMP9 and neutrophils (MPO), accompanied by significant co-localization in the plaques ( Fig. 9 B ) . These results were further confirmed through quantitative analysis which showed positive correlations between MMP9 and MPO, CSF1R and CD86, and CASQ2 and CD206 expression levels ( Fig. 9 C ) . 3.10 TFs and ceRNA regulatory networks for the identified genes Regulatory networks associated with ceRNAs and TFs were constructed to elucidate the underlying mechanisms of ACAS pathogenesis. The results indicated complex interactions governing the expression of the identified genes. The ceRNA theory revealed a competitive binding activity involving lncRNAs (acting as molecular sponges) and the target genes for the miRNA. Notably, we observed that ANPEP interacted with 39 miRNAs versus 32 competing lncRNAs, CASQ2 with 11 miRNAs against 23 lncRNAs, CSF1R with 13 miRNAs versus 40 lncRNAs, and MMP9 with 1 miRNA against 3 lncRNAs ( Fig. 10 ) . The TF network identified multiple TFs associated with the regulation of the target genes, including four TFs predicted to bind to ANPEP, six to CASQ2, four to CSF1R, and seven to MMP9 ( Fig. 11 ) . 4. Discussion The incidence of postoperative stroke in ACAS remains significantly high, ranging between 2–20%, despite significant advancements in the surgical management of this conditio[ 15 ]. Notably, severe adverse consequences have been associated with the middle-aged and elderly populations presenting with ACAS, specifically, patients often suffer from serious complications and significant economic burden that substantially impact their quality of life. Recently, research has been increasing interest in focused on identifying the molecular biomarkers associated with the pathogenesis of ACAS. In this study, we explored the underlying mechanisms of ACAS—which is considered as the major risk factor for IS—by applying bioinformatics to investigate the causative genes associated with ACAS. In our study, four signature genes associated with ACAS were identified: MMP9, CASQ2, ANPEP and CSF1R. Among them, CASQ2 was downregulated, while MMP9, ANPEP, and CSF1R were upregulated. Intergroup differential analysis between the training and validation sets identified four DEGs. Notably, all the four genes exhibited high AUC values through ROC analysis, indicating they had an excellent diagnostic performance. We further validated the diagnostic value of these four identified genes through analysis of atherosclerotic plaque and non-plaque intimal tissue samples obtained from patients with ACAS admitted at our hospital center. Additionally, our analysis explored the infiltrating cell types associated with these four identified genes. Based on our findings, we hypothesized that these identified genes are correlated with the pathogenesis of ACAS and warrant further investigation. The receptor for colony-stimulating factor 1 receptor (CSF1R), also known as the macrophage colony-stimulating factor receptor, is a membrane-spanning tyrosine kinase receptor. This receptor is that is located on the surface of various cell types, including microglia, bone marrow-derived macrophages, and monocytes[ 16 ]. Typically, this receptor occur in an autoinhibited state and is functionally activated upon dimerization, resulting in the auto-phosphorylation of various tyrosine residues. Notably, this process triggers a signaling cascade that subsequently enhances the internalization of the receptor[ 17 ]. Research has shown that inhibition of the CSF1R under physiological condition potentially results in a reversible decrease in population of the microglia[ 18 ]. Additionally, research has revealed that inhibition of CSF1R in microglia significantly reduces their population in vivo ; this observation offers valuable opportunity for future studies involving microglia[ 19 ]. Furthermore, CSF1R has been shown to stimulate ERK1/2-mediated signaling in microglia and activate Akt[ 20 ]. Consequently, it is critical to investigate the role of CSF1R in regulating the M0 macrophages in ACAS and its underlying molecular mechanisms. Developing therapeutic strategies specifically targeting M0 macrophages to promote atherosclerotic plaque stabilization and potentially reverse ACAS, thereby contributing to a reduced risk of stroke. Expression of proteolytic enzymes such as MMPs and cathepsin cysteine proteases (CCPs), along with the depletion of their inhibitors, promotes the process of plaque ulceration and the subsequent rupture of these plaques[ 21 ]. Specifically, MMPs 2, 7, 8, 9, and 13 have been associated with plaque instability. Advanced atherosclerotic lesions exhibit elevated expression of MMPs 2 and 9[ 22 ]. In cases of myocardial infarction, MMP-9 has been shown to be associated with tissue remodeling and rupture of atherosclerotic plaques[ 23 ]. A growing body of evidence indicates that MMP-9—a substance released by activated macrophages and capable of degrading the extracellular matrix—is significantly expressed in vulnerable plaques[ 24 ]. These findings from previous studies further validated the reliability of our identification results. Within the CASQ family, CASQ2 serves as the main calcium-binding reservoir protein, functioning as a calcium sensor and facilitating calcium release into the cytoplasm through the erythropurine receptor 2 channles[ 25 ]. Macrophages, which are the primary immune cells in ACAS, playing key role in its onset, progression, and invasion[ 26 ]. Previous studies have reported that CASQ2 alleviates lung cancer by inhibiting both M2 tumor-associated macrophage polarization and JAK/STAT pathway[ 27 ]. This finding is consistent with our findings. Notably, we found that the expression level of CASQ2 was downregulated in ACAS plaque tissues compared to the adjacent non-plaque intimal tissues; however, this observation warrants further investigation to assess its significance in the context of ACAS pathogenesis. Aminopeptidase N, a membrane-bound zinc-dependent peptidase, plays a key role in the regulation of angiogenesis[ 28 ]. Research has shown that ANPEP deficiency results in the enlargement of atherosclerotic lesions and the increase of necrotic area[ 29 ]. Notably, modulating the intestinal expression of ANPEP and improving the circulating cholesterol distribution can attenuate atherosclerosis[ 30 ]. In this study, we found that the expression of ANPEP was correlated with the pathogenesis of ACAS. Consequently, based on the four identified target genes, we developed a prediction nomogram model for the prevalence of ACAS. This model integrates the four characteristic gene markers to jointly diagnose or predict the pathogenic risk of ACAS, providing each patient with an individualized risk probability score. This approach supports clinical decision-making and advances the implementation of personalized medicine. We performed GSEA, and the results revealed several delineated pathways that are distinctly active in both high and low-expression groups. The high-expression group was mainly enriched in five pathways: cellular receptor signaling pathway, leishmaniasis infection, lysosomes, PPAR signaling pathway, and Toll-like receptor signaling pathway. Among these pathways, the PPAR signaling pathway has been implicated in the regulation of triglycerides, total cholesterol, and free fatty acids. These factors are significant risk factors involved in the pathogenesis of atherosclerosis[ 31 ]. Furthermore, the Toll-like receptor pathway plays a significant role in endothelial dysfunction, immune cell interactions, and various inflammatory processes in the pathogenesis of atherosclerosis[ 32 ]. The lysosomes serve as critical molecules regulating various molecular functions, including lipid degradation, autophagy, apoptosis, inflammasomes, lysosomal biogenesis, and macrophage polarization. Subsequently, lysosomes play significant roles in the occurrence and development of atherosclerotic plaque[ 33 ]. The low-expression group exhibited significant enrichment in multiple pathways—including those associated with dilated cardiomyopathy, pyruvate metabolism, hypertrophic cardiomyopathy, spliceosome, and TGF-β signaling pathway. Notably, the TGF-β signaling pathway is commonly considered a significant contributor to the occurrence of atherosclerosis-associated vascular inflammation. Research has shown that inhibition of the endothelial TGF-β signaling in hyperlipidemic mice reduces inflammation and vascular permeability, subsequently suppressing progression of the condition and inducing regression of established lesions[ 34 ]. Furthermore, pyruvate metabolism is implicated in chronic inflammation, as well as the aberrant proliferation and migration of vascular smooth muscle cells, which is critical in the pathological development and progression of atherosclerotic disease[ 35 ]. Despite significant efforts in elucidating multiple pathways involved in the pathogenesis of atherosclerosis, some key pathways remain largely underexplored, including the spliceosome and leishmaniasis infection. Notably, exploring the involvement of these pathways in atherosclerosis may provide novel therapeutic targets and facilitate the development of more precise and personalized treatment strategies for ACAS. Notably, with the ongoing advancement of sequencing technologies, increasing attention is being directed toward RNA abundance and the functional diversity of TFs. Hence, the regulatory networks involving key genes, ceRNAs, and TFs were constructed to identify novel therapeutic targets for ACAS treatment. A substantial portion of the human genome is transcribed into ncRNAs, which are now recognized as potential biomarkers and therapeutic targets[ 36 ]. Among these, circulating miRNAs are particularly promising due to their critical roles in the pathogenesis of ACAS, particularly through modulation of inflammatory responses and lipid metabolism[ 37 ]. For example, hsa-miR-27a-3p derived from extracellular vesicles promotes M2 macrophage polarization, thereby promoting cellular proliferation and migration[ 38 ]. Similarly, hsa-miR-140-5p can downregulate C-reactive protein expression, a factor closely associated with the formation of atherosclerosis[ 39 ]. MicroRNA-486-5p was identified as a diagnostic marker of ACAS, it mitigates endothelial dysfunction by inhibiting oxidative stress and inflammation[ 40 ]. LncRNAs also play crucial regulatory roles in ACAS. They influence gene expression levels related to endothelial dysfunction, smooth muscle cell proliferation, and macrophage dysfunction in atherosclerotic plaques[ 41 ]. For instance, lncRNA RhabdoMyoSarcoma 2-associated Transcript was upregulated in ACAS patients, and demonstrated high predictive accuracy for ACAS patients[ 42 ]. SNHG14 acts as a sponge for miR-145, thereby regulating cell proliferation involved in restenosis[ 43 ]. Abnormality in lncRNA THRIL expression has been implicated in various disorders correlated with ACAS[ 44 ]. Furthermore, TFs regulate the immune microenvironment by regulating macrophage functions in atherosclerosis through pathways involving cytokine signaling, lipid signaling, and foam cell formation[ 45 ]. For example, downregulated RUNX1 inhibits ox-LDL-induced lipid accumulation and inflammation in macrophages[ 46 ]. Additionally, IFIT1 participates in the inflammatory response triggered by LPS in vascular endothelial cells and is upregulated in aortic plaques of pristane-treated ApoE−/− mice[ 47 ]. Research has also revealed that BNFAT5 can induce vascular endothelial cell apoptosis and inflammatory response, contributing in the pathogenesis of ACAS[ 48 ]. Collectively, these findings support the reliability and biological relevance of the ceRNA and TF regulatory networks constructed in this study, which center around the identified signature genes. Further in-depth exploration of the underlying mechanisms and the functional roles of these ceRNAs and TFs may enhance our understanding of ACAS pathogenesis and aid in the development of precision therapeutic strategies. While this study has presented valuable findings in the context of ACAS, it has several limitations that must be acknowledged. Firstly, by integrating PPI network analysis, MCODE clustering and three ML algorithms, the diagnostic performance of the four signature genes associated with ACAS was verified using an external dataset. However, prospective cohort studies are required to explore their biological significance. Secondly, while this study verified the differential expression of the four signature genes and their association with immune cell infiltration in carotid plaques and adjacent intimal tissues from patients with ACAS at a single center, multicenter studies are needed to enhance the generalizability of these findings. Finally, additional experimental designs are warranted to clarify the potential underlying mechanisms of the four signature genes in the pathogenesis of ACAS. 5. Conclusion In this study, four key pathogenic genes—MMP9, CASQ2, ANPEP, and CSF1R— associated with the pathogenesis of ACAS were identified through an integrated approach involving PPI network analysis and MCODE clustering, combined with three ML algorithms. Additionally, the corresponding ceRNAs and TFs regulating these genes were predicted. These predicted genes and their regulatory elements may serve as novel diagnostic biomarkers and potential therapeutic targets for ACAS. Abbreviations Atherosclerotic carotid artery stenosis (ACAS) Ischemic stroke (IS) Machine learning (ML) Protein-protein interaction (PPI) Molecular complex detection (MCODE) Least Absolute Shrinkage and Selection Operator (LASSO) Random Forest (RF) Support Vector Machine Recursive Feature Elimination (SVM-RFE) National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) Principal Component Analysis (PCA) Carotid endarterectomy (CEA) Differentially expressed genes (DEGs) Gene Ontology (GO) Kyoto Encyclopedia of Genes and Genomes (KEGG) Gene set enrichment analysis (GSEA) Decision curve analysis (DCA) Reverse transcription quantitative polymerase chain reaction (RT-qPCR) Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) Immunofluorescence (IF) Competing endogenous RNA (ceRNA) Transcription factors (TFs) Long non-coding RNA (lncRNA) Colony-stimulating factor-1 receptor (CSF1R) C-C chemokine receptor (CCR1) Matrix metalloproteinase (MMP) Declarations Ethics approval This study was approved by the Ethics Committee of Renmin Hospital of Wuhan university and The First Affiliated Hospital of Yangtze University. Competing interests The authors declare that there are no competing interests or conflicts of interests relating to this work. Funding This work was supported by the Scientific Research Project of Hubei Health Commission (Grant No. WJ2023M079) and Natural Science Foundation of Hubei (Grant No. 2024AFB707). Author Contribution All of the authors have significantly contributed to this manuscript and are in agreement with its content. WZ and LWK conducted conceptualization and methodology. HPD was involved in both drafting of the original manuscript and generation of visual elements of the study. LJJ and YL were involving in the editing of the final manuscript as well as performing data analysis. All authors have read and agreed to the final version of the manuscript. Acknowledgement The authors would like to express their sincere gratitude to the Scientific research project of Hubei Health Commission for the financial support. Data Availability The raw data supporting the findings of this study are available from the corresponding author upon reasonable request and without undue reservation. References Campbell BCV, De Silva DA, Macleod MR, Coutts SB, Schwamm LH and Davis SM, et al. Ischaemic stroke. Nat Rev Dis Primers. 2019; 5(1): 70. Saini V, Guada L and Yavagal DR. Global epidemiology of stroke and access to acute ischemic stroke interventions. Neurology. 2021; 97(20 Suppl 2): S6-16. Bonati LH, Gregson J, Dobson J, Mccabe DJH, Nederkoorn PJ and van der Worp HB, et al. Restenosis and risk of stroke after stenting or endarterectomy for symptomatic carotid stenosis in the international carotid stenting study (ICSS): secondary analysis of a randomised trial. Lancet Neurol. 2018; 17(7): 587–96. Golnari P, Nazari P, Hurley MC, Shaibani A, Ansari SA and Potts MB, et al. Continued dominance of carotid endarterectomy over stenting in the united states: volumes, outcomes, and complications from the national inpatient sample (1997–2015). World Neurosurg. 2022; 163: e238-52. Gaba K, Ringleb PA and Halliday A. Asymptomatic carotid stenosis: intervention or best medical therapy? Curr Neurol Neurosci Rep. 2018; 18(11): 80. White CJ, Brott TG, Gray WA, Heck D, Jovin T and Lyden SP, et al. Carotid artery stenting: JACC state-of-the-art review. J Am Coll Cardiol. 2022; 80(2): 155–70. Swanson K, Wu E, Zhang A, Alizadeh AA and Zou J. From patterns to patients: advances in clinical machine learning for cancer diagnosis, prognosis, and treatment. Cell. 2023; 186(8): 1772–91. Qin X, Yi S, Rong J, Lu H, Ji B and Zhang W, et al. Identification of anoikis-related genes classification patterns and immune infiltration characterization in ischemic stroke based on machine learning. Front Aging Neurosci. 2023; 15: 1142163. Gene ontology consortium: going forward. Nucleic Acids Res. 2015; 43(Database issue): D1049-56. Kanehisa M and Goto S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 2000; 28(1): 27–30. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL and Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005; 102(43): 15545–50. Bader GD and Hogue CWV. An automated method for finding molecular complexes in large protein interaction networks. BMC Bioinformatics. 2003; 4: 2. Furio-Tari P, Tarazona S, Gabaldon T, Enright AJ and Conesa A. Spongescan: a web for detecting microRNA binding elements in lncRNA sequences. Nucleic Acids Res. 2016; 44(W1): W176-80. Zhou G, Soufan O, Ewald J, Hancock REW, Basu N and Xia J. NetworkAnalyst 3.0: a visual analytics platform for comprehensive gene expression profiling and meta-analysis. Nucleic Acids Res. 2019; 47(W1): W234-41. Maravic-Stojkovic V, Lausevic-Vuk LJ, Obradovic M, Jovanovic P, Tanaskovic S and Stojkovic B, et al. Copeptin level after carotid endarterectomy and perioperative stroke. Angiology. 2014; 65(2): 122–9. Barca C, Foray C, Hermann S, Herrlinger U, Remory I and Laoui D, et al. The colony stimulating factor-1 receptor (CSF-1r)-mediated regulation of microglia/macrophages as a target for neurological disorders (glioma, stroke). Front Immunol. 2021; 12: 787307. Gutmann DH and Kettenmann H. Microglia/brain macrophages as central drivers of brain tumor pathobiology. Neuron. 2019; 104(3): 442–9. Spangenberg EE, Lee RJ, Najafi AR, Rice RA, Elmore MRP and Blurton-Jones M, et al. Eliminating microglia in alzheimer's mice prevents neuronal loss without modulating amyloid-beta pathology. Brain. 2016; 139(Pt 4): 1265–81. Ginhoux F, Greter M, Leboeuf M, Nandi S, See P and Gokhan S, et al. Fate mapping analysis reveals that adult microglia derive from primitive macrophages. Science. 2010; 330(6005): 841–5. Otxoa-De-Amezaga A, Miro-Mur F, Pedragosa J, Gallizioli M, Justicia C and Gaja-Capdevila N, et al. Microglial cell loss after ischemic stroke favors brain neutrophil accumulation. Acta Neuropathol. 2019; 137(2): 321–41. Wijeratne T, Menon R, Sales C, Karimi L and Crewther S. Carotid artery stenosis and inflammatory biomarkers: the role of inflammation-induced immunological responses affecting the vascular systems. Ann Transl Med. 2020; 8(19): 1276. Liu B, Su L, Loo SJ, Gao Y, Khin E and Kong X, et al. Matrix metallopeptidase 9 contributes to the beginning of plaque and is a potential biomarker for the early identification of atherosclerosis in asymptomatic patients with diabetes. Front Endocrinol (Lausanne). 2024; 15: 16. Somuncu MU, Pusuroglu H, Karakurt H, Bolat I, Karakurt ST and Demir AR, et al. The prognostic value of elevated matrix metalloproteinase-9 in patients undergoing primary percutaneous coronary intervention for ST-elevation myocardial infarction: a two-year prospective study. Rev Port Cardiol. 2020; 39(5): 267–76. Camare C, Pucelle M, Negre-Salvayre A and Salvayre R. Angiogenesis in the atherosclerotic plaque. Redox Biol. 2017; 12: 18–34. Chen H, Valle G, Furlan S, Nani A, Gyorke S and Fill M, et al. Mechanism of calsequestrin regulation of single cardiac ryanodine receptor in normal and pathological conditions. J Gen Physiol. 2013; 142(2): 127–36. Xiao Q, Hou R, Xie L, Niu M, Pan X and Zhu X. Macrophage metabolic reprogramming and atherosclerotic plaque microenvironment: fostering each other? Clin Transl Med. 2023; 13(5): e1257. Ding Y, Yuan X, Wang Y and Yan J. CASQ2 alleviates lung cancer by inhibiting m2 tumor-associated macrophage polarization and JAK/STAT pathway. J Biochem Mol Toxicol. 2024; 38(8): e23801. Shui L, Meng Y, Huang C, Qian Y and Liu J. Aminopeptidase n expression in the endometrium could affect endometrial receptivity. Biochem Biophys Res Commun. 2019; 514(2): 469–74. Devarakonda CV, Pereira FE, Smith JD, Shapiro LH and Ghosh M. CD13 deficiency leads to increased oxidative stress and larger atherosclerotic lesions. Atherosclerosis. 2019; 287: 70–80. Du H, Yang L, Zhang H, Zhang X and Shao H. LncRNA TUG1 silencing enhances proliferation and migration of ox-LDL-treated human umbilical vein endothelial cells and promotes atherosclerotic vascular injury repairing via the runx2/ANPEP axis. Int J Cardiol. 2021; 338: 204–14. Miao M, Wang X, Liu T, Li Y, Yu W and Yang T, et al. Targeting PPARs for therapy of atherosclerosis: a review. Int J Biol Macromol. 2023; 242(Pt 2): 125008. Jin M, Fang J, Wang J, Shao X, Xu S and Liu P, et al. Regulation of toll-like receptor (TLR) signaling pathways in atherosclerosis: from mechanisms to targeted therapeutics. Acta Pharmacol Sin. 2023; 44(12): 2358–75. Zhang Z, Yue P, Lu T, Wang Y, Wei Y and Wei X. Role of lysosomes in physiological activities, diseases, and therapy. J Hematol Oncol. 2021; 14(1): 79. Chen P, Qin L, Li G, Wang Z, Dahlman JE and Malagon-Lopez J, et al. Endothelial TGF-beta signalling drives vascular inflammation and atherosclerosis. Nat Metab. 2019; 1(9): 912–26. Li L, Wang M, Ma Q, Ye J and Sun G. Role of glycolysis in the development of atherosclerosis. Am J Physiol Cell Physiol. 2022; 323(2): C617-29. Chen H, Shi Z, Feng H, Wang R, Zhang Y and Xie J, et al. The relationship between dizziness and cervical artery stenosis. Neuroreport. 2015; 26(18): 1112–8. Liu H, Zhou J, Jiang W and Wang F. Analysis of the diagnostic and prognostic value of mir-9-5p in carotid artery stenosis. Bosn J Basic Med Sci. 2021; 21(6): 724–9. Zhao G, Yu H, Ding L, Wang W, Wang H and Hu Y, et al. MicroRNA-27a-3p delivered by extracellular vesicles from glioblastoma cells induces m2 macrophage polarization via the EZH1/KDM3a/CTGF axis. Cell Death Discov. 2022; 8(1): 260. Teng L and Meng R. Long non-coding RNA MALAT1 promotes acute cerebral infarction through miRNAs-mediated hs-CRP regulation. J Mol Neurosci. 2019; 69(3): 494–504. Zhu B, Liu W, Xu Q and Liu H. MicroRNA-486-5p functions as a diagnostic marker for carotid artery stenosis and prevents endothelial dysfunction through inhibiting inflammation and oxidative stress. Bioengineered. 2022; 13(4): 8667–75. Villa C and Stoccoro A. Epigenetic peripheral biomarkers for early diagnosis of alzheimer's disease. Genes (Basel). 2022; 13(8). Wang C, Zhao F, He Y, E Y and Li S. Long non-coding RNA RMST serves as a diagnostic biomarker in patients with carotid artery stenosis and predicts the occurrence of cerebral ischemic event: a retrospective study. Vascular. 2023; 31(5): 908–13. Ma H and Dong A. Dysregulation of lncRNA SNHG1/mir-145 axis affects the biological function of human carotid artery smooth muscle cells as a mechanism of carotid artery restenosis. Exp Ther Med. 2021; 21(5): 423. Wang F, Wang Y and Guo X. LncRNA THRIL functions as a marker for carotid artery stenosis and affects the biological function of human aortic endothelial cell. J Inflamm Res. 2023; 16: 2437–46. Hao Z, Zhang Y, Li Y, Zhao J, Zhou Y and Qiu J, et al. The association between ideal cardiovascular health metrics and extracranial carotid artery stenosis in a northern chinese population: a cross-sectional study. Sci Rep. 2016; 6: 7. Yan Y, Lu Y, Liu Y, Zhang J, Wang S and Dong J, et al. Identification of circular RNA hsa_circ_0034621 as a novel biomarker for carotid atherosclerosis and the potential function as a regulator of NLRP3 inflammasome. Atherosclerosis. 2024; 391: 10. Zhang M, Zhu Y, Zhu J, Xie Y, Wu R and Zhong J, et al. Circ_0086296 induced atherosclerotic lesions via the IFIT1 / STAT1 feedback loop by sponging mir-576-3p. Cell Mol Biol Lett. 2022; 27(1): 26. Xie X, Huang C, Xu D, Liu Y, Hu M and Long J, et al. Elevation of hypertonicity - induced protein NFAT5 promotes apoptosis of human umbilical vein endothelial cells through the NF - κb pathway. Mol Med Rep. 2021; 23(3): 8. Additional Declarations No competing interests reported. Supplementary Files SupplementaryTable1.docx Cite Share Download PDF Status: Published Journal Publication published 04 Nov, 2025 Read the published version in European Journal of Medical Research → Version 1 posted Editorial decision: Revision requested 17 Sep, 2025 Reviews received at journal 26 Aug, 2025 Reviews received at journal 23 Aug, 2025 Reviews received at journal 13 Aug, 2025 Reviewers agreed at journal 08 Aug, 2025 Reviewers agreed at journal 06 Aug, 2025 Reviewers agreed at journal 06 Aug, 2025 Reviewers invited by journal 06 Aug, 2025 Editor assigned by journal 16 Jul, 2025 Submission checks completed at journal 15 Jul, 2025 First submitted to journal 13 Jul, 2025 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-7112702\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":false,\"archivedVersions\":[],\"articleType\":\"Research Article\",\"associatedPublications\":[],\"authors\":[{\"id\":497697470,\"identity\":\"36845455-d113-460b-a1a0-07dd93a64cbd\",\"order_by\":0,\"name\":\"Lei Yang\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"The First Affiliated Hospital of Yangtze University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Lei\",\"middleName\":\"\",\"lastName\":\"Yang\",\"suffix\":\"\"},{\"id\":497697471,\"identity\":\"423638f3-858e-43da-bebd-df8fd1180eb5\",\"order_by\":1,\"name\":\"Peidong He\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Renmin Hospital of Wuhan University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Peidong\",\"middleName\":\"\",\"lastName\":\"He\",\"suffix\":\"\"},{\"id\":497697472,\"identity\":\"7326501c-4f30-4d64-b041-369faba8b88f\",\"order_by\":2,\"name\":\"Wenkang Lei\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Renmin Hospital of Wuhan University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Wenkang\",\"middleName\":\"\",\"lastName\":\"Lei\",\"suffix\":\"\"},{\"id\":497697473,\"identity\":\"2c33512b-3bbb-4703-a8dc-4121624688e2\",\"order_by\":3,\"name\":\"Zhen Wang\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Renmin Hospital of Wuhan University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Zhen\",\"middleName\":\"\",\"lastName\":\"Wang\",\"suffix\":\"\"},{\"id\":497697474,\"identity\":\"dd0f8155-16b0-4295-904d-e154e683ec2d\",\"order_by\":4,\"name\":\"Jingjing Liang\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4klEQVRIiWNgGAWjYBACA/bG9s9/Kmrq7fubDxz48IMYLTyHjzHwnDmWYCBxLPHgzB5itEikpTHwtjAnGDDkGB/mYCNCi7lEjtkDyQa2PHOGMx8OM/AwyPOLHcCvxbLnjbmB4Q6ZYsvm3g2HCywYDGfOTiDgsOM5BhKJZ9gYGw6c3XB4Bg9DgsFtQloOALUcbGMGasl5cJiHjRgtJ9LSJBvbmBM3HMhhIFLLmcOHjRnOHDOWnHHMABjIEsT4pbHxMUNFjRw/f/PjDx9+2MjzSxPQgg4kSFM+CkbBKBgFowA7AACyJE9QArGDRQAAAABJRU5ErkJggg==\",\"orcid\":\"\",\"institution\":\"Renmin Hospital of Wuhan University\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Jingjing\",\"middleName\":\"\",\"lastName\":\"Liang\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2025-07-13 10:38:03\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-7112702/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-7112702/v1\",\"draftVersion\":[],\"editorialEvents\":[{\"content\":\"https://doi.org/10.1186/s40001-025-03330-8\",\"type\":\"published\",\"date\":\"2025-11-04T15:57:38+00:00\"}],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":88804265,\"identity\":\"7a4557e6-577b-4c4a-bee2-88b1912666d4\",\"added_by\":\"auto\",\"created_at\":\"2025-08-11 14:41:33\",\"extension\":\"png\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":2838065,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eIdentification of the DEGs. \\u003c/strong\\u003e(A) Tissue samples from the four datasets exhibited batch effects. (B) The batch effects in the tissue samples from the fourdatasets were eliminated after batch correction. (C) Volcano plots depicting all genes. (D) Heatmaps illustrating the 52 downregulated and 33 upregulated genes.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"Figure1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/48ce815d33ff1ea034052112.png\"},{\"id\":88804267,\"identity\":\"ccc45939-0f96-4736-a924-07ead3a0e98b\",\"added_by\":\"auto\",\"created_at\":\"2025-08-11 14:41:33\",\"extension\":\"png\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":2924010,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eFunctional enrichment analysis of the 85 DEGs.\\u003c/strong\\u003e(A) GO analysis. (B) KEGG analysis. (C) Five active pathways identified in the low-expression group. (D) Five active pathways identified in the high-expression group.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"Figure2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/a2d288851434cd231570de3f.png\"},{\"id\":88804798,\"identity\":\"37faa0e0-54cc-4867-afa3-924fd705bb48\",\"added_by\":\"auto\",\"created_at\":\"2025-08-11 14:49:33\",\"extension\":\"png\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":8443839,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003ePPI network and MCODE analysis. \\u003c/strong\\u003e(A) PPI network of the DEGs. (B-D) Three key functional modulesidentified using MCODE analysis. Blue represents downregulated genes and red represents upregulated genes.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"Figure3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/c7284588f06a17832d43e951.png\"},{\"id\":88804800,\"identity\":\"ed2bf8cc-4234-4809-bd3e-69adb4fa4931\",\"added_by\":\"auto\",\"created_at\":\"2025-08-11 14:49:33\",\"extension\":\"png\",\"order_by\":4,\"title\":\"Figure 4\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":3282872,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eThree machine learning algorithms of the feature genes. \\u003c/strong\\u003e(A) LASSO coefficient path diagram with each curve representing one gene. (B) LASSO regression analysis cross validation curve. The LASSO fit was the best and the cross-validation error was smallest when 9genes were used in the analysis. (C) The highest accuracy (0.746) with SVM-RFE algorithm when there were 11 genes. (D) The lowest error rate with SVM-RFE method when there are 11 genes (0.254). (E) Relationship between the number of RF trees and the cabinet of error rates. The red dotted line indicates the error rate when identifying the control group, the green line signifies the error rate when identifying the experimental group, and the black dotted line denotes the combined error rate. (F) Genes are listed in descending order of importance. (G) Venn plots of the genes gained with the three ML algorithms.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"Figure4.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/78d24a1d8a3ab0fbe11b062f.png\"},{\"id\":88804268,\"identity\":\"cb2d4fbd-7271-4a9e-b2fe-e81bf332d08b\",\"added_by\":\"auto\",\"created_at\":\"2025-08-11 14:41:33\",\"extension\":\"png\",\"order_by\":5,\"title\":\"Figure 5\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":3946591,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eIntergroup difference analysis of the feature genes. \\u003c/strong\\u003e(A) Violin plots demonstrating the results of expression difference analysis for four feature genes in the training group. (B) Violin plots demonstrating the results of expression difference analysis for four feature genes in the validation group. (C) Line plots illustrating the expression levels of the 4 feature genes.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"Figure5.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/91892e87be79bcc09bf46524.png\"},{\"id\":88806043,\"identity\":\"3ee862ac-57c4-409a-9641-bc41cb29d511\",\"added_by\":\"auto\",\"created_at\":\"2025-08-11 14:57:33\",\"extension\":\"png\",\"order_by\":6,\"title\":\"Figure 6\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":2236898,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eROC analysis of the feature genes. \\u003c/strong\\u003e(A) ROC analysis of the identified genes in the training group. (B) ROC analysis of the identified genes in the validation group\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"Figure6.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/bfaefd33c6fb5653ae78b66a.png\"},{\"id\":88804804,\"identity\":\"8dee92e9-c258-4e06-ba2c-7c2ecc30bb71\",\"added_by\":\"auto\",\"created_at\":\"2025-08-11 14:49:33\",\"extension\":\"png\",\"order_by\":7,\"title\":\"Figure 7\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":1403290,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eAlignment diagram model of carotid artery stenosis patients. \\u003c/strong\\u003e(A) Prediction of ACAS in the alignment diagram. (B) The predictive accuracy of the modelwith calibration curve. (C) Clinical impact curve ofthe model. (D) The clinical benefit for patients with ACAS using decision curve analyses.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"Figure7.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/477e952eeda10019362aaea7.png\"},{\"id\":88804801,\"identity\":\"b53ff3c0-f24c-4fb7-9498-ef3c6ef4b1d1\",\"added_by\":\"auto\",\"created_at\":\"2025-08-11 14:49:33\",\"extension\":\"png\",\"order_by\":8,\"title\":\"Figure 8\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":3014957,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eImmune-related functions and immune cell correlation analysis of characterized genes.\\u003c/strong\\u003e (A–D) Box line plots showing the differences in immune-related functions between low- and high-expression groups for ANPEP (A), CASQ2 (B), CSF1R (C), and MMP9 (D), respectively. (E–H) Lollipop charts illustrating the correlation between ANPEP (E), CASQ2 (F), CSF1R (G), and MMP9 (H) with the 22 immune cells, respectively.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"Figure8.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/d3330b3bb1661c9448aa9118.png\"},{\"id\":88806044,\"identity\":\"60c66ec1-ed7c-42f6-8a6d-4a1176f4ef0b\",\"added_by\":\"auto\",\"created_at\":\"2025-08-11 14:57:33\",\"extension\":\"png\",\"order_by\":9,\"title\":\"Figure 9\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":2169454,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eValidation of the characterized genes with clinical samples. \\u003c/strong\\u003e(A) Validation of the characterized geneswith RT-qPCR assays. (n=15), * \\u003cem\\u003ep \\u003c/em\\u003e\\u0026lt; 0.05, **\\u003cem\\u003ep \\u003c/em\\u003e\\u0026lt; 0.01. (B) IF double-labeling of plaque tissues with characterized genes. Scale bar=50um. (C) Correlation plot of characterized genes and marker protein expression.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"Figure9.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/01a60d53da8778a37f174429.png\"},{\"id\":88808130,\"identity\":\"89381515-712e-44b3-b60b-51342fe9e3f7\",\"added_by\":\"auto\",\"created_at\":\"2025-08-11 15:13:33\",\"extension\":\"png\",\"order_by\":10,\"title\":\"Figure 10\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":1699637,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eCeRNA networks of feature genes.\\u003c/strong\\u003e The ceRNA regulatory networks for ANPEP, CASQ2, CSF1R and MMP9. Blue represents feature genes, orange represents miRNA, green represents lncRNA.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"Figure10.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/54fa71d50c3baca8ec33c8fd.png\"},{\"id\":88806046,\"identity\":\"b8a80b81-63c9-417c-a295-4c01d1656cdc\",\"added_by\":\"auto\",\"created_at\":\"2025-08-11 14:57:33\",\"extension\":\"png\",\"order_by\":11,\"title\":\"Figure 11\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":2123930,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eTFs regulatory networks of feature genes:\\u003c/strong\\u003e TF regulatory networks of ANPEP, CASQ2, CSF1R and MMP9.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"Figure11.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/6fb35c0ffe9f2721e28c06e9.png\"},{\"id\":95564199,\"identity\":\"e7b1496b-0dbd-48f9-a07b-1f01e0cddfab\",\"added_by\":\"auto\",\"created_at\":\"2025-11-10 16:08:49\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":32800372,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/d29f7e7d-a6da-4330-913b-9edeaecbfa30.pdf\"},{\"id\":88807007,\"identity\":\"0126ff8d-9a97-4f8a-8cc3-e888de198092\",\"added_by\":\"auto\",\"created_at\":\"2025-08-11 15:05:33\",\"extension\":\"docx\",\"order_by\":2,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":15726,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"SupplementaryTable1.docx\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-7112702/v1/cf710f1016d203a18ee00ee8.docx\"}],\"financialInterests\":\"No competing interests reported.\",\"formattedTitle\":\"Identification of key genes and regulatory networks associated with atherosclerotic carotid artery stenosis through comprehensive bioinformatics analysis and machine learning\",\"fulltext\":[{\"header\":\"1. Introduction\",\"content\":\"\\u003cp\\u003eAtherosclerotic carotid artery stenosis (ACAS), which is the narrowing of the carotid artery due to the formation of atherosclerotic plaque, is a highly prevalent condition. This condition may progress from partial stenosis to complete occlusion, leading to adverse events such as ischemic stroke (IS) or transient ischemic attack[\\u003cspan citationid=\\\"CR1\\\" class=\\\"CitationRef\\\"\\u003e1\\u003c/span\\u003e]. The prevalence of ACAS is increasing due to population aging and increased incidences of cerebrovascular risk factors. Associated stroke events have imposed a significant economic burden on both the families and the broader healthcare system[\\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e]. Consequently, there is an urgent need to investigate its causative factors, to facilitate the implementation of preventive and control measures to improve the prognosis of ACAS patients.\\u003c/p\\u003e\\u003cp\\u003ePatients with ACAS are at increased risk of developing stroke. Typically, most of them require carotid endarterectomy (CEA) or stenting to improve cerebral blood flow, however, these therapies are not effective in completely treating ACAS, with many patients subsequently developing restenosis[\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e]. Notably, CEA is associated with approximately 2\\u0026ndash;20% incidence of postoperative stroke\\u0026mdash;a rare and severe complication[\\u003cspan citationid=\\\"CR4\\\" class=\\\"CitationRef\\\"\\u003e4\\u003c/span\\u003e]. While existing pharmacological agents are primarily used to manage blood pressure and atherothrombosis, the genetic and molecular mechanisms underlying ACAS remain incompletely understood[\\u003cspan citationid=\\\"CR5\\\" class=\\\"CitationRef\\\"\\u003e5\\u003c/span\\u003e]. Consequently, in-depth exploration of these molecular mechanisms offer a promising potential for the development of novel therapeutic strategies for ACAS.\\u003c/p\\u003e\\u003cp\\u003eNotably, ACAS is a regulatory condition that arises from key inflammatory, immune, and metabolic pathways. Research has shown that reversing the pathogenesis of ACAS, along with improving prognostic outcomes, has been a challenging task[\\u003cspan citationid=\\\"CR6\\\" class=\\\"CitationRef\\\"\\u003e6\\u003c/span\\u003e]. However, machine learning (ML) offers a promising avenue towards this goal. Notably, ML is characterized by reduced error rates, with superior predictive performance, thereby enhancing its widespread adoption for target screening, prognosis prediction, and providing guidance for optimal therapeutic option[\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR8\\\" class=\\\"CitationRef\\\"\\u003e8\\u003c/span\\u003e]. This study used an integrated approach involving protein-protein interaction (PPI) network analysis and molecular complex detection (MCODE) to explore the molecular relationships among key genes involved in ACAS. Machine learning models, including least absolute shrinkage and selection operator (LASSO), random forest (RF), and support vector machine recursive feature elimination (SVM-RFE), were utilized for a comprehensive analysis to identify critical targets. These identified targets may provide a valuable therapeutic foundation for developing precise risk prediction models and personalized treatment decisions in ACAS.\\u003c/p\\u003e\"},{\"header\":\"2. Materials and methods\",\"content\":\"\\u003cdiv id=\\\"Sec3\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e2.1 Retrieval and merging of datasets\\u003c/h2\\u003e\\u003cp\\u003eThe ACAS datasets used in this study were retrieved from the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) database and they included the following: GSE100927 (12 normal endothelium samples\\u0026thinsp;+\\u0026thinsp;29 carotid atherosclerosis samples), GSE28829 (13 early plaque samples\\u0026thinsp;+\\u0026thinsp;16 advanced carotid atherosclerosis samples), GSE198600 (5 stable carotid atherosclerotic plaque samples\\u0026thinsp;+\\u0026thinsp;6 unstable carotid atherosclerosis samples), GSE43292 (32 normal endothelium samples\\u0026thinsp;+\\u0026thinsp;32 carotid atherosclerosis samples), and GSE163154 (16 stable carotid atherosclerosis samples\\u0026thinsp;+\\u0026thinsp;27 unstable carotid atherosclerosis samples). Notably, the first four datasets were merged to form the training set (n\\u0026thinsp;=\\u0026thinsp;145), with GSE163154 (n\\u0026thinsp;=\\u0026thinsp;43) serving as the validation set. Details of the datasets are listed in Table\\u0026nbsp;\\u003cspan refid=\\\"Tab2\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e. Batch correction was performed on the datasets using the \\u0026ldquo;sva\\u0026rdquo; package (version 3.50.0), while the principal component analysis (PCA) was used to visualize the results before and after correction.\\u003c/p\\u003e\\u003cp\\u003e\\u003cdiv class=\\\"gridtable\\\"\\u003e\\u003ctable float=\\\"Yes\\\" id=\\\"Tab1\\\" border=\\\"1\\\"\\u003e\\u003ccaption language=\\\"En\\\"\\u003e\\u003cdiv class=\\\"CaptionNumber\\\"\\u003eTable 1\\u003c/div\\u003e\\u003cdiv class=\\\"CaptionContent\\\"\\u003e\\u003cp\\u003eBasic information on the GEO datasets used in the study.\\u003c/p\\u003e\\u003c/div\\u003e\\u003c/caption\\u003e\\u003ccolgroup cols=\\\"6\\\"\\u003e\\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c1\\\" colnum=\\\"1\\\"\\u003e\\u003c/div\\u003e\\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c2\\\" colnum=\\\"2\\\"\\u003e\\u003c/div\\u003e\\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c3\\\" colnum=\\\"3\\\"\\u003e\\u003c/div\\u003e\\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c4\\\" colnum=\\\"4\\\"\\u003e\\u003c/div\\u003e\\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c5\\\" colnum=\\\"5\\\"\\u003e\\u003c/div\\u003e\\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c6\\\" colnum=\\\"6\\\"\\u003e\\u003c/div\\u003e\\u003cthead\\u003e\\u003ctr\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eID\\u003c/p\\u003e\\u003c/th\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003eGSE series\\u003c/p\\u003e\\u003c/th\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003eDisease\\u003c/p\\u003e\\u003c/th\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003eSamples\\u003c/p\\u003e\\u003c/th\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003ePlatform\\u003c/p\\u003e\\u003c/th\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c6\\\"\\u003e\\u003cp\\u003eGroup\\u003c/p\\u003e\\u003c/th\\u003e\\u003c/tr\\u003e\\u003c/thead\\u003e\\u003ctbody\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e1\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003eGSE43292\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003eCAS\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e32 carotid atherosclerosis plaques and 32 endosomal \\u003c/p\\u003e\\u003cp\\u003etissues\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003eGPL6244\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c6\\\"\\u003e\\u003cp\\u003eDiscovery set\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e2\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003eGSE100927\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003eCAS\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e29 carotid atherosclerosis plaques and 12 endosomal \\u003c/p\\u003e\\u003cp\\u003etissues\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003eGPL17077\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c6\\\"\\u003e\\u003cp\\u003eDiscovery set\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e3\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003eGSE198600\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003eCAS\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e6 unstable carotid plaque samples and 5 stable carotid plaque samples\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003eGPL11154\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c6\\\"\\u003e\\u003cp\\u003eDiscovery set\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e4\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003eGSE28829\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003eCAS\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e16 advanced carotid plaque samples and 13 early carotid plaque samples\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003eGPL570\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c6\\\"\\u003e\\u003cp\\u003eDiscovery set\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e5\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003eGSE163154\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003eCAS\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003e27 unstable plaque samples and 16 stable plaque samples\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e\\u003cp\\u003eGPL6104\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c6\\\"\\u003e\\u003cp\\u003evalidation set\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003c/tbody\\u003e\\u003c/colgroup\\u003e\\u003c/table\\u003e\\u003c/div\\u003e\\u003c/p\\u003e\\u003cdiv id=\\\"Sec4\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e2.2 Clinical samples of ACAS\\u003c/h2\\u003e\\u003cp\\u003eA total of 20 ACAS patients admitted at Renmin Hospital of Wuhan University and 10 ACAS patients admitted at The First Affiliated Hospital of Yangtze University between 2021 and 2023 were enrolled in this study. Both the control and experimental groups included 15 ACAS patients who underwent CEA. The experimental group comprised the atherosclerotic plaque tissues, whereas the control group comprised the adjacent non-plaque intimal tissues. The atherosclerotic plaque and adjacent non-plaque intimal tissues were collected after the CEA. The study protocol was approved by the Clinical Research Ethics Committee of Renmin Hospital of Wuhan University (Approval No. WDRY2023-K123) and the ethics committee of the First Affiliated Hospital of Yangtze University (Approval No. LL202301). Additionally, the study methodologies strictly adhere to the relevant guidelines and regulations. All participants signed the informed consent form.\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec5\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e2.3 Identification of differentially expressed genes (DEGs)\\u003c/h2\\u003e\\u003cp\\u003eThe \\u0026ldquo;limma\\u0026rdquo; package (version 3.60.4) was used to identify the DEGs by performing normalization and analysis of gene expression patterns. The criteria for identifying the DEGs were based on an adjusted p-value\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05 and an absolute log\\u003csub\\u003e2\\u003c/sub\\u003efold change (|logFC|)\\u0026thinsp;\\u0026gt;\\u0026thinsp;1. The results were visualized by generating heatmaps using the \\u0026ldquo;pheatmap\\u0026rdquo; R package (version 1.0.12).\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec6\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e2.4 Functional enrichment analysis\\u003c/h2\\u003e\\u003cp\\u003eFunctional and pathway enrichment analyses of the DEGs were performed with the \\u0026ldquo;ClusterProfiler,\\u0026rdquo; \\u0026ldquo;enrichplot,\\u0026rdquo; and \\u0026ldquo;org.Hs.eg.db\\u0026rdquo; packages. Gene ontology (GO)[\\u003cspan citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e] and Kyoto encyclopedia of genes and genomes (KEGG)[\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e] were employed for detailed annotation and pathway enrichment analyses. The reference genome file \\u0026ldquo;c2.cp.kegg.Hs.symbols.gmt\\u0026rdquo; was used to conduct the gene set enrichment analysis (GSEA)[\\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e] to distinguish the differences in the associated pathways. All results were visualized using the \\u0026ldquo;ggplot2\\u0026rdquo; R package (version 3.5.1).\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec7\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e2.5 Protein-protein interaction (PPI) and MCODE analysis\\u003c/h2\\u003e\\u003cp\\u003eThe PPI network of the 85 DEGs was constructed by uploading them into the STRING database (\\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttp://string-db.org\\u003c/span\\u003e\\u003cspan address=\\\"http://string-db.org\\\" targettype=\\\"URL\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e) with a medium confidence level of 0.400. The PPI network was visualized using the Cytoscape_3.10.0 (version 3.10.0) software. To identify key sub-networks and core genes, MCODE[\\u003cspan citationid=\\\"CR12\\\" class=\\\"CitationRef\\\"\\u003e12\\u003c/span\\u003e], analysis was conducted within Cytoscape, facilitating the extraction of critical sub-networks and core genes based on the topological relationships between edges and nodes within the PPI network. Furthermore, functional module extraction and node information analysis were performed using MCODE with specific parameter settings as follows: Degree Cutoff of 2, Node Score Cutoff of 0.2, K-Core of 2, and a Max Depth from Seed of 100.\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec8\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e2.6 Three Machine learning methods for filtering feature genes\\u003c/h2\\u003e\\u003cp\\u003ePathogenic genes were identified from the functional modules using three approaches: LASSO, SVM-RFE, and Random Forest. Notably, the LASSO regression was used to generate coefficient paths and cross-validation curves to determine optimal λ values and characteristic gene numbers. The SVM-RFE evaluation plots illustrated the performance of feature selection based on 10-fold cross-validation accuracy and error rates. The RF algorithm identified optimal tree numbers (500 trees) via error minimization, with top-ranked genes selected based on importance scores for further analysis.\\u003c/p\\u003e\\u003cp\\u003eThe overlapping results from the three algorithms were identified, and a Venn diagram was constructed using the \\u0026ldquo;Venn Diagram\\u0026rdquo; R package (version 1.7.3). Furthermore, ROC curves were plotted, and then the area under the curve (AUC) was calculated with the \\u0026ldquo;pROC\\u0026rdquo; and \\u0026ldquo;InpROC\\u0026rdquo; packages in R software (version 1.18.5) to determine the predictive value of feature genes in both the training and validation sets.\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec9\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e2.7 Construction of carotid stenosis maps\\u003c/h2\\u003e\\u003cp\\u003eA nomogram was constructed based on the identified feature genes using the \\u0026ldquo;rms\\u0026rdquo; (version 6.8) and \\u0026ldquo;rmda\\u0026rdquo; (version 1.6) packages in R. Visual assessment of the predictive performance of the nomogram was evaluated by generating calibration curves. Clinical impact of the nomogram was further assessed by generating clinical impact curves, which illustrate the potential clinical applicability of the model. Furthermore, the clinical utility and net benefit of the nomogram was assed using the decision curve analysis (DCA), thereby ensuring its clinical relevance and value in practical decision-making scenarios.\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec10\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e2.8 Immune-related function and immune cell correlation analysis\\u003c/h2\\u003e\\u003cp\\u003eA total of 83 carotid atherosclerosis samples were stratified into high-expression groups based on the expression levels of the target genes. The differences in immune-related functions between the two groups were performed through functional differential analysis using the R packages \\u0026ldquo;GSVA\\u0026rdquo; (version 1.52.3) and \\u0026ldquo;GSEABase\\u0026rdquo; (version 1.66.0). Furthermore, correlation analysis was conducted to investigate the associations between feature genes and immune cells, using the \\u0026ldquo;ggpubr\\u0026rdquo; (version 0.6.0), \\u0026ldquo;reshape2\\u0026rdquo; (version 1.4.4), and \\u0026lsquo;ggExtra\\u0026rdquo; (version 0.10.1) packages for visualization and comprehensive analysis.\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec11\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e2.9 Reverse transcription quantitative polymerase chain reaction (RT-qPCR)\\u003c/h2\\u003e\\u003cp\\u003eThe tissue samples were lysed using the Trizol reagent. Subsequently, RevertAidTM First Strand cDNA Synthesis Kit (Takara, Japan) was used to generate complementary DNA with the following reaction conditions: exposure to room temperature for 10 mins, 42\\u0026deg;C for 60 mins, 95\\u0026deg;C for 5 mins, and freezing for 5 mins. Using SYBR Green reagent, the RT-qPCR reaction was conducted using the 7500 PCR equipment. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal reference to normalize target gene expression levels, which were quantified using the 2\\u003csup\\u003e\\u0026minus;ΔΔct\\u003c/sup\\u003e technique. The primer sequences used are shown in \\u003cb\\u003eTable \\u003cspan refid=\\\"MOESM1\\\" class=\\\"InternalRef\\\"\\u003eS1\\u003c/span\\u003e\\u003c/b\\u003e.\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec12\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e\\u003cb\\u003e2.10 Immunofluorescence (IF) staining\\u003c/b\\u003e\\u003c/h2\\u003e\\u003cp\\u003eThe tissue samples were fixed in 10% formalin, serially dehydrated in ethanol at increasing percentages, and then incubated in chloroform. Subsequently, the fixed tissues were embedded in paraffin for double immunofluorescent labeling. Following antigen retrieval using hot citrate buffer, the deparaffinized sections were then soaked in a blocking reagent for 1 hour. The primary antibodies were incubated with human carotid atherosclerotic plaques at 4\\u0026deg;C overnight. The primary antibodies used included: CD68 (servicebio, GB113150), CD206 (servicebio, GB115273), CASQ2 (servicebio, GB114847), CSF1R (servicebio, GB11581), MPO (servicebio, GB11224), and MMP9 (servicebio, GB12132). The tissue sections were thoroughly washed in readiness for secondary antibody incubation, which involved 2 hours of incubation with Alexa Fluor secondary antibodies from Thermo Fisher Scientific. The stained tissue sections were then mounted using Vectashield Mounting Medium (Burlingame, CA, USA) to preserve fluorescence and ensure high-quality visualization. Furthermore, tissue sections were treated with 2 mol/L HCl for 20 mins at 30\\u0026deg;C to denature DNA for BrdU staining. Subsequently, the sections were rinsed with PBS and incubated with 0.1 mol/L borate buffer for 10 mins at room temperature to neutralize the acid. Following an additional PBS wash, the slices were subjected to the previously described immunofluorescence staining procedure.\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec13\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e2.11 Construction of regulatory network involving characteristic genes, competing endogenous RNA (ceRNA), and transcription factors (TFs)\\u003c/h2\\u003e\\u003cp\\u003eTo identify miRNAs potentially interacting with the feature genes, miRNA prediction was performed using miRDB, miRanda, and TargetScan techniques. Subsequent analysis involved only miRNAs commonly identified by all three tools. Long non-coding RNA (lncRNA) interacting with the identified miRNAs were analyzed using the SpongeScan database to identify potential miRNA-binding lncRNAs[\\u003cspan citationid=\\\"CR13\\\" class=\\\"CitationRef\\\"\\u003e13\\u003c/span\\u003e]. Subsequently, a ceRNA regulatory network was constructed based on these interaction results and visualized using the Cytoscape software, integrating the interactions between lncRNAs, miRNAs, and feature genes. Additionally, a gene-TF regulatory network was constructed using the NetworkAnalyst (\\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttp://www.networkanalyst.ca\\u003c/span\\u003e\\u003cspan address=\\\"http://www.networkanalyst.ca\\\" targettype=\\\"URL\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e)[\\u003cspan citationid=\\\"CR14\\\" class=\\\"CitationRef\\\"\\u003e14\\u003c/span\\u003e], thereby providing a comprehensive analysis visualization of the regulatory mechanisms associated with the feature genes.\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec14\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e2.12 Statistical analysis\\u003c/h2\\u003e\\u003cp\\u003eAll statistical analyses were performed using the R software (version 4.4.0). Normally distributed variables were analyzed using the t-test, while the Wilcoxon test was used for non-normally distributed variables. Correlation analyses were conducted using the Pearson analysis for linear relationships, while the Spearman analysis was used for monotonic relationships. All statistical tests were two-sided, with an adjusted \\u003cem\\u003ep\\u003c/em\\u003e-value\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05 considered to be statistically significant.\\u003c/p\\u003e\\u003c/div\\u003e\"},{\"header\":\"3. Results\",\"content\":\"\\u003cdiv id=\\\"Sec16\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e3.1 A total of 85 DEGs were identified\\u003c/h2\\u003e\\u003cp\\u003eBatch correction was performed before any differential analysis. Notably, the PCA analysis revealed that the samples from different datasets were distinct before the correction, indicating that there was a batch effect between these samples (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003eA). However, after batch correction, these samples were distributed randomly, indicating elimination of the batch effect (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003eB). All gene volcanoes were then plotted (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003eC), and differential analysis identified 85 DEGs, with 33 downregulated and 52 upregulated genes (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003eD).\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec17\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e3.2 Functional and pathway analysis of the 85 DEGs\\u003c/h2\\u003e\\u003cp\\u003eFunctional enrichment analysis was conducted on the 85 DEGs. Notably, the KEGG and GO analysis revealed that these genes were mainly associated with negative regulation of transport activities, myeloid leukocyte differentiation, and leukocyte chemotaxis (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eA), as well as pathways such as phagolysosomes, tuberculosis, diabetic cardiomyopathy, lipids and atherosclerosis (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eB). Furthermore, GSEA analysis was conducted to identify the differences in the pathways associated with the high and low-expressed genes. The results revealed that dilated cardiomyopathy, pyruvate metabolism, hypertrophic cardiomyopathy, spliceosome, and TGF-β signaling pathways were highly activated in the low-expressed genes (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eC), whereas cellular receptor signaling pathway, leishmaniasis infection, lysosome, PPAR-signaling pathway, and Toll -like receptor signaling pathway were significantly activated in the high-expressed genes (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eD).\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec18\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e3.3 Three important functional modules contained 26 genes using MCODE analysis\\u003c/h2\\u003e\\u003cp\\u003eThe PPI map of the 85 DEGS was constructed (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eA). To investigate the potential molecular mechanisms underlying ACAS, a modular network analysis was performed using the MCODE algorithm to identify densely connected regions within the PPI network. This analysis revealed three key modules comprising a total of 26 genes, representing potential core therapeutic targets for ACAS (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eB-D). The detailed information of each module is shown in \\u003cb\\u003eTable\\u0026nbsp;2\\u003c/b\\u003e.\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003c/div\\u003e\\u003cp\\u003e\\u003cdiv class=\\\"gridtable\\\"\\u003e\\u003ctable float=\\\"Yes\\\" id=\\\"Tab2\\\" border=\\\"1\\\"\\u003e\\u003ccaption language=\\\"En\\\"\\u003e\\u003cdiv class=\\\"CaptionNumber\\\"\\u003eTable 2\\u003c/div\\u003e\\u003cdiv class=\\\"CaptionContent\\\"\\u003e\\u003cp\\u003eThe results of the MCODE analysis.\\u003c/p\\u003e\\u003c/div\\u003e\\u003c/caption\\u003e\\u003ccolgroup cols=\\\"4\\\"\\u003e\\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c1\\\" colnum=\\\"1\\\"\\u003e\\u003c/div\\u003e\\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c2\\\" colnum=\\\"2\\\"\\u003e\\u003c/div\\u003e\\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c3\\\" colnum=\\\"3\\\"\\u003e\\u003c/div\\u003e\\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c4\\\" colnum=\\\"4\\\"\\u003e\\u003c/div\\u003e\\u003cthead\\u003e\\u003ctr\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003eCluster\\u003c/p\\u003e\\u003c/th\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003eNodes\\u003c/p\\u003e\\u003c/th\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003eEdges\\u003c/p\\u003e\\u003c/th\\u003e\\u003cth align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003eNode IDs\\u003c/p\\u003e\\u003c/th\\u003e\\u003c/tr\\u003e\\u003c/thead\\u003e\\u003ctbody\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e1\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e18\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e101\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003eTREM1, ITGAM, MMP9, CCR1, LY86, C1QB, CTSS, CD52, ITGB2, SPP1, NCF2, CD36, APOE, CD14, CSF1R, LAPTM5, ANPEP, TYROBP\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e2\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e5\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e9\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003eMYOCD, ATP1A2, PLN, RYR2, CASQ2\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003ctr\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e\\u003cp\\u003e3\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e\\u003cp\\u003e3\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c3\\\"\\u003e\\u003cp\\u003e3\\u003c/p\\u003e\\u003c/td\\u003e\\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e\\u003cp\\u003eMMP12, MMP1, CHI3L1\\u003c/p\\u003e\\u003c/td\\u003e\\u003c/tr\\u003e\\u003c/tbody\\u003e\\u003c/colgroup\\u003e\\u003c/table\\u003e\\u003c/div\\u003e\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec19\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e3.4 Four feature genes were identified using the three ML algorithms\\u003c/h2\\u003e\\u003cp\\u003eThe LASSO algorithm identified a total of 9 feature genes\\u0026mdash;including MMP9, CCR1, CD14, CSF1R, ANPEP, MYOCD, ATP1A2, CASQ2, and CHI3L1 (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig4\\\" class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eA, B). The SVM-RFE analysis on the 26 genes identified 11 feature genes\\u0026mdash;including PLN, ANPEP, CSF1R, CTSS, MYOCD, C1QB, APOE, CHI3L1, MMP9, ITGAM, and CASQ2 (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig4\\\" class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eC, D). Additionally, the RF method identified the top 10 feature genes based on relative importance, including ITGB2, APOE, MMP1, SPP1, MMP9, CASQ2, CSF1R, C1QB, NCF2, and ANPEP (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig4\\\" class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eE, F). Subsequently, the intersection of the identified genes was identified as four common feature genes: ANPEP, CASQ2, CSF1R, and MMP9 (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig4\\\" class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eG).\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec20\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e3.5 Intergroup differences among the four feature genes\\u003c/h2\\u003e\\u003cp\\u003eTo visualize the differences in the expression levels among the four identified genes, line and violin plots were generated. The results revealed significant differences between the expression of the four feature genes in both the experimental and control groups (p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.01, Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig5\\\" class=\\\"InternalRef\\\"\\u003e5\\u003c/span\\u003eA, B). Additionally, ANPEP, CSF1R, and MMP9 were upregulated, while CASQ2 was downregulated in the experimental group (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig5\\\" class=\\\"InternalRef\\\"\\u003e5\\u003c/span\\u003eC).\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec21\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e3.6 ROC analysis of the 4 signature genes\\u003c/h2\\u003e\\u003cp\\u003eTo verify the diagnostic accuracy of the identified genes, ROC analysis was conducted. In the training set, the AUC values for ANPEP, CASQ2, CSF1R, and MMP9 were 0.851, 0.864, 0.914, and 0.861, respectively (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig6\\\" class=\\\"InternalRef\\\"\\u003e6\\u003c/span\\u003eA). In the validation set, the AUC values for ANPEP, CASQ2, CSF1R, and MMP9 were 0.947, 0.921, 0.944 and 0.935, respectively (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig6\\\" class=\\\"InternalRef\\\"\\u003e6\\u003c/span\\u003eB).\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec22\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e3.7 Association between the prevalence of ACAS and characteristic genes\\u003c/h2\\u003e\\u003cp\\u003eA column chart was constructed to illustrate the diagnostic performance of the combined expression levels of the four signature genes in identifying ACAS (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig7\\\" class=\\\"InternalRef\\\"\\u003e7\\u003c/span\\u003eA). The calibration curve demonstrated high accuracy in predicting the disease prevalence \\u003cb\\u003e(\\u003c/b\\u003eFig.\\u0026nbsp;\\u003cspan refid=\\\"Fig7\\\" class=\\\"InternalRef\\\"\\u003e7\\u003c/span\\u003eB\\u003cb\\u003e)\\u003c/b\\u003e, while the clinical impact curve validate that the model had a robust predictive capability across various thresholds \\u003cb\\u003e(\\u003c/b\\u003eFig.\\u0026nbsp;\\u003cspan refid=\\\"Fig7\\\" class=\\\"InternalRef\\\"\\u003e7\\u003c/span\\u003eC\\u003cb\\u003e)\\u003c/b\\u003e. Furthermore, DCA indicated significant clinical benefits for patients diagnosed with ACAS using the nomogram. This indicated that the nomogram could be utilized as a reliable diagnostic tool in clinical practice \\u003cb\\u003e(\\u003c/b\\u003eFig.\\u0026nbsp;\\u003cspan refid=\\\"Fig7\\\" class=\\\"InternalRef\\\"\\u003e7\\u003c/span\\u003eD\\u003cb\\u003e)\\u003c/b\\u003e.\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec23\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e3.8 Immune-related function and immune cell correlation analysis of the characteristic genes\\u003c/h2\\u003e\\u003cp\\u003eThere were significant differences in immune functionality between the high- and low-expression groups as evidenced by the immune-related function and immune cell correlation analysis of the identified genes \\u003cb\\u003e(\\u003c/b\\u003eFig.\\u0026nbsp;\\u003cspan refid=\\\"Fig8\\\" class=\\\"InternalRef\\\"\\u003e8\\u003c/span\\u003eA-D\\u003cb\\u003e)\\u003c/b\\u003e. This observation indicates that the identified genes significantly impact the immune microenvironment. Notably, ANPEP was positively correlated with immune-suppressive cells, such as Tregs and resting NK cells, while negatively correlated with antigen-presenting and cytotoxic cells, such as CD8\\u0026thinsp;+\\u0026thinsp;T and dendritic cells \\u003cb\\u003e(\\u003c/b\\u003eFig.\\u0026nbsp;\\u003cspan refid=\\\"Fig8\\\" class=\\\"InternalRef\\\"\\u003e8\\u003c/span\\u003eE\\u003cb\\u003e)\\u003c/b\\u003e. Additionally, CASQ2 influenced macrophage polarization, revealing high propensity towards the anti-inflammatory M2 macrophages over pro-inflammatory M0 macrophages \\u003cb\\u003e(\\u003c/b\\u003eFig.\\u0026nbsp;\\u003cspan refid=\\\"Fig8\\\" class=\\\"InternalRef\\\"\\u003e8\\u003c/span\\u003eF\\u003cb\\u003e)\\u003c/b\\u003e. Also, CSF1R modulated macrophage activity, positively correlating with M0 macrophages and inversely with activated NK cells and M2 macrophages \\u003cb\\u003e(\\u003c/b\\u003eFig.\\u0026nbsp;\\u003cspan refid=\\\"Fig8\\\" class=\\\"InternalRef\\\"\\u003e8\\u003c/span\\u003eG\\u003cb\\u003e)\\u003c/b\\u003e. The MMP9 exhibited associations with neutrophil and resting NK cell activity, while inversely correlated with activated CD8\\u0026thinsp;+\\u0026thinsp;T cells and NK cells \\u003cb\\u003e(\\u003c/b\\u003eFig.\\u0026nbsp;\\u003cspan refid=\\\"Fig8\\\" class=\\\"InternalRef\\\"\\u003e8\\u003c/span\\u003eH\\u003cb\\u003e)\\u003c/b\\u003e, suggesting a role in immune surveillance and inflammation. Collectively, these findings highlight the complex interactions between gene expression patterns and immune cell dynamics, providing valuable insights into their contributions to the modulation of the immune microenvironment.\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec24\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e3.9 Validation of the identified genes using tissue samples\\u003c/h2\\u003e\\u003cp\\u003eThe expression levels of the identified genes were assessed between the carotid atherosclerotic plaque tissues and adjacent non-plaque intimal tissues. Notably, RT-qPCR analysis revealed that the levels of MMP9, ANPEP, and CSF1R were upregulated, while CASQ2 was downregulated (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig9\\\" class=\\\"InternalRef\\\"\\u003e9\\u003c/span\\u003eA). The immune cell correlations among the identified genes in ACAS were evaluated using the IF double-labeling technique on carotid atherosclerotic plaques from patients undergoing CEA. The following results were observed: positive correlations between CASQ2 and M2 macrophages (CD206), CSF1R and M1 macrophages (CD86), and MMP9 and neutrophils (MPO), accompanied by significant co-localization in the plaques \\u003cb\\u003e(\\u003c/b\\u003eFig.\\u0026nbsp;\\u003cspan refid=\\\"Fig9\\\" class=\\\"InternalRef\\\"\\u003e9\\u003c/span\\u003eB\\u003cb\\u003e)\\u003c/b\\u003e. These results were further confirmed through quantitative analysis which showed positive correlations between MMP9 and MPO, CSF1R and CD86, and CASQ2 and CD206 expression levels \\u003cb\\u003e(\\u003c/b\\u003eFig.\\u0026nbsp;\\u003cspan refid=\\\"Fig9\\\" class=\\\"InternalRef\\\"\\u003e9\\u003c/span\\u003eC\\u003cb\\u003e)\\u003c/b\\u003e.\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003c/div\\u003e\\u003cdiv id=\\\"Sec25\\\" class=\\\"Section2\\\"\\u003e\\u003ch2\\u003e3.10 TFs and ceRNA regulatory networks for the identified genes\\u003c/h2\\u003e\\u003cp\\u003eRegulatory networks associated with ceRNAs and TFs were constructed to elucidate the underlying mechanisms of ACAS pathogenesis. The results indicated complex interactions governing the expression of the identified genes. The ceRNA theory revealed a competitive binding activity involving lncRNAs (acting as molecular sponges) and the target genes for the miRNA. Notably, we observed that ANPEP interacted with 39 miRNAs versus 32 competing lncRNAs, CASQ2 with 11 miRNAs against 23 lncRNAs, CSF1R with 13 miRNAs versus 40 lncRNAs, and MMP9 with 1 miRNA against 3 lncRNAs \\u003cb\\u003e(\\u003c/b\\u003eFig.\\u0026nbsp;\\u003cspan refid=\\\"Fig10\\\" class=\\\"InternalRef\\\"\\u003e10\\u003c/span\\u003e\\u003cb\\u003e)\\u003c/b\\u003e. The TF network identified multiple TFs associated with the regulation of the target genes, including four TFs predicted to bind to ANPEP, six to CASQ2, four to CSF1R, and seven to MMP9 \\u003cb\\u003e(\\u003c/b\\u003eFig.\\u0026nbsp;\\u003cspan refid=\\\"Fig11\\\" class=\\\"InternalRef\\\"\\u003e11\\u003c/span\\u003e\\u003cb\\u003e)\\u003c/b\\u003e.\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003c/div\\u003e\"},{\"header\":\"4. Discussion\",\"content\":\"\\u003cp\\u003eThe incidence of postoperative stroke in ACAS remains significantly high, ranging between 2\\u0026ndash;20%, despite significant advancements in the surgical management of this conditio[\\u003cspan citationid=\\\"CR15\\\" class=\\\"CitationRef\\\"\\u003e15\\u003c/span\\u003e]. Notably, severe adverse consequences have been associated with the middle-aged and elderly populations presenting with ACAS, specifically, patients often suffer from serious complications and significant economic burden that substantially impact their quality of life. Recently, research has been increasing interest in focused on identifying the molecular biomarkers associated with the pathogenesis of ACAS. In this study, we explored the underlying mechanisms of ACAS\\u0026mdash;which is considered as the major risk factor for IS\\u0026mdash;by applying bioinformatics to investigate the causative genes associated with ACAS.\\u003c/p\\u003e\\u003cp\\u003eIn our study, four signature genes associated with ACAS were identified: MMP9, CASQ2, ANPEP and CSF1R. Among them, CASQ2 was downregulated, while MMP9, ANPEP, and CSF1R were upregulated. Intergroup differential analysis between the training and validation sets identified four DEGs. Notably, all the four genes exhibited high AUC values through ROC analysis, indicating they had an excellent diagnostic performance. We further validated the diagnostic value of these four identified genes through analysis of atherosclerotic plaque and non-plaque intimal tissue samples obtained from patients with ACAS admitted at our hospital center. Additionally, our analysis explored the infiltrating cell types associated with these four identified genes. Based on our findings, we hypothesized that these identified genes are correlated with the pathogenesis of ACAS and warrant further investigation.\\u003c/p\\u003e\\u003cp\\u003eThe receptor for colony-stimulating factor 1 receptor (CSF1R), also known as the macrophage colony-stimulating factor receptor, is a membrane-spanning tyrosine kinase receptor. This receptor is that is located on the surface of various cell types, including microglia, bone marrow-derived macrophages, and monocytes[\\u003cspan citationid=\\\"CR16\\\" class=\\\"CitationRef\\\"\\u003e16\\u003c/span\\u003e]. Typically, this receptor occur in an autoinhibited state and is functionally activated upon dimerization, resulting in the auto-phosphorylation of various tyrosine residues. Notably, this process triggers a signaling cascade that subsequently enhances the internalization of the receptor[\\u003cspan citationid=\\\"CR17\\\" class=\\\"CitationRef\\\"\\u003e17\\u003c/span\\u003e]. Research has shown that inhibition of the CSF1R under physiological condition potentially results in a reversible decrease in population of the microglia[\\u003cspan citationid=\\\"CR18\\\" class=\\\"CitationRef\\\"\\u003e18\\u003c/span\\u003e]. Additionally, research has revealed that inhibition of CSF1R in microglia significantly reduces their population \\u003cem\\u003ein vivo\\u003c/em\\u003e; this observation offers valuable opportunity for future studies involving microglia[\\u003cspan citationid=\\\"CR19\\\" class=\\\"CitationRef\\\"\\u003e19\\u003c/span\\u003e]. Furthermore, CSF1R has been shown to stimulate ERK1/2-mediated signaling in microglia and activate Akt[\\u003cspan citationid=\\\"CR20\\\" class=\\\"CitationRef\\\"\\u003e20\\u003c/span\\u003e]. Consequently, it is critical to investigate the role of CSF1R in regulating the M0 macrophages in ACAS and its underlying molecular mechanisms. Developing therapeutic strategies specifically targeting M0 macrophages to promote atherosclerotic plaque stabilization and potentially reverse ACAS, thereby contributing to a reduced risk of stroke.\\u003c/p\\u003e\\u003cp\\u003eExpression of proteolytic enzymes such as MMPs and cathepsin cysteine proteases (CCPs), along with the depletion of their inhibitors, promotes the process of plaque ulceration and the subsequent rupture of these plaques[\\u003cspan citationid=\\\"CR21\\\" class=\\\"CitationRef\\\"\\u003e21\\u003c/span\\u003e]. Specifically, MMPs 2, 7, 8, 9, and 13 have been associated with plaque instability. Advanced atherosclerotic lesions exhibit elevated expression of MMPs 2 and 9[\\u003cspan citationid=\\\"CR22\\\" class=\\\"CitationRef\\\"\\u003e22\\u003c/span\\u003e]. In cases of myocardial infarction, MMP-9 has been shown to be associated with tissue remodeling and rupture of atherosclerotic plaques[\\u003cspan citationid=\\\"CR23\\\" class=\\\"CitationRef\\\"\\u003e23\\u003c/span\\u003e]. A growing body of evidence indicates that MMP-9\\u0026mdash;a substance released by activated macrophages and capable of degrading the extracellular matrix\\u0026mdash;is significantly expressed in vulnerable plaques[\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e]. These findings from previous studies further validated the reliability of our identification results. Within the CASQ family, CASQ2 serves as the main calcium-binding reservoir protein, functioning as a calcium sensor and facilitating calcium release into the cytoplasm through the erythropurine receptor 2 channles[\\u003cspan citationid=\\\"CR25\\\" class=\\\"CitationRef\\\"\\u003e25\\u003c/span\\u003e]. Macrophages, which are the primary immune cells in ACAS, playing key role in its onset, progression, and invasion[\\u003cspan citationid=\\\"CR26\\\" class=\\\"CitationRef\\\"\\u003e26\\u003c/span\\u003e]. Previous studies have reported that CASQ2 alleviates lung cancer by inhibiting both M2 tumor-associated macrophage polarization and JAK/STAT pathway[\\u003cspan citationid=\\\"CR27\\\" class=\\\"CitationRef\\\"\\u003e27\\u003c/span\\u003e]. This finding is consistent with our findings. Notably, we found that the expression level of CASQ2 was downregulated in ACAS plaque tissues compared to the adjacent non-plaque intimal tissues; however, this observation warrants further investigation to assess its significance in the context of ACAS pathogenesis. Aminopeptidase N, a membrane-bound zinc-dependent peptidase, plays a key role in the regulation of angiogenesis[\\u003cspan citationid=\\\"CR28\\\" class=\\\"CitationRef\\\"\\u003e28\\u003c/span\\u003e]. Research has shown that ANPEP deficiency results in the enlargement of atherosclerotic lesions and the increase of necrotic area[\\u003cspan citationid=\\\"CR29\\\" class=\\\"CitationRef\\\"\\u003e29\\u003c/span\\u003e]. Notably, modulating the intestinal expression of ANPEP and improving the circulating cholesterol distribution can attenuate atherosclerosis[\\u003cspan citationid=\\\"CR30\\\" class=\\\"CitationRef\\\"\\u003e30\\u003c/span\\u003e]. In this study, we found that the expression of ANPEP was correlated with the pathogenesis of ACAS. Consequently, based on the four identified target genes, we developed a prediction nomogram model for the prevalence of ACAS. This model integrates the four characteristic gene markers to jointly diagnose or predict the pathogenic risk of ACAS, providing each patient with an individualized risk probability score. This approach supports clinical decision-making and advances the implementation of personalized medicine.\\u003c/p\\u003e\\u003cp\\u003eWe performed GSEA, and the results revealed several delineated pathways that are distinctly active in both high and low-expression groups. The high-expression group was mainly enriched in five pathways: cellular receptor signaling pathway, leishmaniasis infection, lysosomes, PPAR signaling pathway, and Toll-like receptor signaling pathway. Among these pathways, the PPAR signaling pathway has been implicated in the regulation of triglycerides, total cholesterol, and free fatty acids. These factors are significant risk factors involved in the pathogenesis of atherosclerosis[\\u003cspan citationid=\\\"CR31\\\" class=\\\"CitationRef\\\"\\u003e31\\u003c/span\\u003e]. Furthermore, the Toll-like receptor pathway plays a significant role in endothelial dysfunction, immune cell interactions, and various inflammatory processes in the pathogenesis of atherosclerosis[\\u003cspan citationid=\\\"CR32\\\" class=\\\"CitationRef\\\"\\u003e32\\u003c/span\\u003e]. The lysosomes serve as critical molecules regulating various molecular functions, including lipid degradation, autophagy, apoptosis, inflammasomes, lysosomal biogenesis, and macrophage polarization. Subsequently, lysosomes play significant roles in the occurrence and development of atherosclerotic plaque[\\u003cspan citationid=\\\"CR33\\\" class=\\\"CitationRef\\\"\\u003e33\\u003c/span\\u003e]. The low-expression group exhibited significant enrichment in multiple pathways\\u0026mdash;including those associated with dilated cardiomyopathy, pyruvate metabolism, hypertrophic cardiomyopathy, spliceosome, and TGF-β signaling pathway. Notably, the TGF-β signaling pathway is commonly considered a significant contributor to the occurrence of atherosclerosis-associated vascular inflammation. Research has shown that inhibition of the endothelial TGF-β signaling in hyperlipidemic mice reduces inflammation and vascular permeability, subsequently suppressing progression of the condition and inducing regression of established lesions[\\u003cspan citationid=\\\"CR34\\\" class=\\\"CitationRef\\\"\\u003e34\\u003c/span\\u003e]. Furthermore, pyruvate metabolism is implicated in chronic inflammation, as well as the aberrant proliferation and migration of vascular smooth muscle cells, which is critical in the pathological development and progression of atherosclerotic disease[\\u003cspan citationid=\\\"CR35\\\" class=\\\"CitationRef\\\"\\u003e35\\u003c/span\\u003e]. Despite significant efforts in elucidating multiple pathways involved in the pathogenesis of atherosclerosis, some key pathways remain largely underexplored, including the spliceosome and leishmaniasis infection. Notably, exploring the involvement of these pathways in atherosclerosis may provide novel therapeutic targets and facilitate the development of more precise and personalized treatment strategies for ACAS.\\u003c/p\\u003e\\u003cp\\u003eNotably, with the ongoing advancement of sequencing technologies, increasing attention is being directed toward RNA abundance and the functional diversity of TFs. Hence, the regulatory networks involving key genes, ceRNAs, and TFs were constructed to identify novel therapeutic targets for ACAS treatment.\\u003c/p\\u003e\\u003cp\\u003eA substantial portion of the human genome is transcribed into ncRNAs, which are now recognized as potential biomarkers and therapeutic targets[\\u003cspan citationid=\\\"CR36\\\" class=\\\"CitationRef\\\"\\u003e36\\u003c/span\\u003e]. Among these, circulating miRNAs are particularly promising due to their critical roles in the pathogenesis of ACAS, particularly through modulation of inflammatory responses and lipid metabolism[\\u003cspan citationid=\\\"CR37\\\" class=\\\"CitationRef\\\"\\u003e37\\u003c/span\\u003e]. For example, hsa-miR-27a-3p derived from extracellular vesicles promotes M2 macrophage polarization, thereby promoting cellular proliferation and migration[\\u003cspan citationid=\\\"CR38\\\" class=\\\"CitationRef\\\"\\u003e38\\u003c/span\\u003e]. Similarly, hsa-miR-140-5p can downregulate C-reactive protein expression, a factor closely associated with the formation of atherosclerosis[\\u003cspan citationid=\\\"CR39\\\" class=\\\"CitationRef\\\"\\u003e39\\u003c/span\\u003e]. MicroRNA-486-5p was identified as a diagnostic marker of ACAS, it mitigates endothelial dysfunction by inhibiting oxidative stress and inflammation[\\u003cspan citationid=\\\"CR40\\\" class=\\\"CitationRef\\\"\\u003e40\\u003c/span\\u003e]. LncRNAs also play crucial regulatory roles in ACAS. They influence gene expression levels related to endothelial dysfunction, smooth muscle cell proliferation, and macrophage dysfunction in atherosclerotic plaques[\\u003cspan citationid=\\\"CR41\\\" class=\\\"CitationRef\\\"\\u003e41\\u003c/span\\u003e]. For instance, lncRNA RhabdoMyoSarcoma 2-associated Transcript was upregulated in ACAS patients, and demonstrated high predictive accuracy for ACAS patients[\\u003cspan citationid=\\\"CR42\\\" class=\\\"CitationRef\\\"\\u003e42\\u003c/span\\u003e]. SNHG14 acts as a sponge for miR-145, thereby regulating cell proliferation involved in restenosis[\\u003cspan citationid=\\\"CR43\\\" class=\\\"CitationRef\\\"\\u003e43\\u003c/span\\u003e]. Abnormality in lncRNA THRIL expression has been implicated in various disorders correlated with ACAS[\\u003cspan citationid=\\\"CR44\\\" class=\\\"CitationRef\\\"\\u003e44\\u003c/span\\u003e]. Furthermore, TFs regulate the immune microenvironment by regulating macrophage functions in atherosclerosis through pathways involving cytokine signaling, lipid signaling, and foam cell formation[\\u003cspan citationid=\\\"CR45\\\" class=\\\"CitationRef\\\"\\u003e45\\u003c/span\\u003e]. For example, downregulated RUNX1 inhibits ox-LDL-induced lipid accumulation and inflammation in macrophages[\\u003cspan citationid=\\\"CR46\\\" class=\\\"CitationRef\\\"\\u003e46\\u003c/span\\u003e]. Additionally, IFIT1 participates in the inflammatory response triggered by LPS in vascular endothelial cells and is upregulated in aortic plaques of pristane-treated ApoE\\u0026minus;/\\u0026minus; mice[\\u003cspan citationid=\\\"CR47\\\" class=\\\"CitationRef\\\"\\u003e47\\u003c/span\\u003e]. Research has also revealed that BNFAT5 can induce vascular endothelial cell apoptosis and inflammatory response, contributing in the pathogenesis of ACAS[\\u003cspan citationid=\\\"CR48\\\" class=\\\"CitationRef\\\"\\u003e48\\u003c/span\\u003e]. Collectively, these findings support the reliability and biological relevance of the ceRNA and TF regulatory networks constructed in this study, which center around the identified signature genes. Further in-depth exploration of the underlying mechanisms and the functional roles of these ceRNAs and TFs may enhance our understanding of ACAS pathogenesis and aid in the development of precision therapeutic strategies.\\u003c/p\\u003e\\u003cp\\u003eWhile this study has presented valuable findings in the context of ACAS, it has several limitations that must be acknowledged. Firstly, by integrating PPI network analysis, MCODE clustering and three ML algorithms, the diagnostic performance of the four signature genes associated with ACAS was verified using an external dataset. However, prospective cohort studies are required to explore their biological significance. Secondly, while this study verified the differential expression of the four signature genes and their association with immune cell infiltration in carotid plaques and adjacent intimal tissues from patients with ACAS at a single center, multicenter studies are needed to enhance the generalizability of these findings. Finally, additional experimental designs are warranted to clarify the potential underlying mechanisms of the four signature genes in the pathogenesis of ACAS.\\u003c/p\\u003e\"},{\"header\":\"5. Conclusion\",\"content\":\"\\u003cp\\u003eIn this study, four key pathogenic genes\\u0026mdash;MMP9, CASQ2, ANPEP, and CSF1R\\u0026mdash; associated with the pathogenesis of ACAS were identified through an integrated approach involving PPI network analysis and MCODE clustering, combined with three ML algorithms. Additionally, the corresponding ceRNAs and TFs regulating these genes were predicted. These predicted genes and their regulatory elements may serve as novel diagnostic biomarkers and potential therapeutic targets for ACAS.\\u003c/p\\u003e\"},{\"header\":\"Abbreviations\",\"content\":\"\\u003cp\\u003eAtherosclerotic carotid artery stenosis (ACAS)\\u003c/p\\u003e\\u003cp\\u003eIschemic stroke (IS)\\u003c/p\\u003e\\u003cp\\u003eMachine learning (ML)\\u003c/p\\u003e\\u003cp\\u003eProtein-protein interaction (PPI)\\u003c/p\\u003e\\u003cp\\u003eMolecular complex detection (MCODE)\\u003c/p\\u003e\\u003cp\\u003eLeast Absolute Shrinkage and Selection Operator (LASSO)\\u003c/p\\u003e\\u003cp\\u003eRandom Forest (RF)\\u003c/p\\u003e\\u003cp\\u003eSupport Vector Machine Recursive Feature Elimination (SVM-RFE)\\u003c/p\\u003e\\u003cp\\u003eNational Center for Biotechnology Information (NCBI)\\u003c/p\\u003e\\u003cp\\u003eGene Expression Omnibus (GEO)\\u003c/p\\u003e\\u003cp\\u003ePrincipal Component Analysis (PCA)\\u003c/p\\u003e\\u003cp\\u003eCarotid endarterectomy (CEA)\\u003c/p\\u003e\\u003cp\\u003eDifferentially expressed genes (DEGs)\\u003c/p\\u003e\\u003cp\\u003eGene Ontology (GO)\\u003c/p\\u003e\\u003cp\\u003eKyoto Encyclopedia of Genes and Genomes (KEGG)\\u003c/p\\u003e\\u003cp\\u003eGene set enrichment analysis (GSEA)\\u003c/p\\u003e\\u003cp\\u003eDecision curve analysis (DCA)\\u003c/p\\u003e\\u003cp\\u003eReverse transcription quantitative polymerase chain reaction (RT-qPCR)\\u003c/p\\u003e\\u003cp\\u003eGlyceraldehyde 3-phosphate dehydrogenase (GAPDH)\\u003c/p\\u003e\\u003cp\\u003eImmunofluorescence (IF)\\u003c/p\\u003e\\u003cp\\u003eCompeting endogenous RNA (ceRNA)\\u003c/p\\u003e\\u003cp\\u003eTranscription factors (TFs)\\u003c/p\\u003e\\u003cp\\u003eLong non-coding RNA (lncRNA)\\u003c/p\\u003e\\u003cp\\u003eColony-stimulating factor-1 receptor (CSF1R)\\u003c/p\\u003e\\u003cp\\u003eC-C chemokine receptor (CCR1)\\u003c/p\\u003e\\u003cp\\u003eMatrix metalloproteinase (MMP)\\u003c/p\\u003e\"},{\"header\":\"Declarations\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eEthics approval\\u003c/strong\\u003e\\u003c/p\\u003e\\u003cp\\u003e This study was approved by the Ethics Committee of Renmin Hospital of Wuhan university and The First Affiliated Hospital of Yangtze University.\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003cp\\u003e\\u003cstrong\\u003eCompeting interests\\u003c/strong\\u003e\\u003c/p\\u003e\\u003cp\\u003eThe authors declare that there are no competing interests or conflicts of interests relating to this work.\\u003c/p\\u003e\\u003cp\\u003e\\u003c/p\\u003e\\u003ch2\\u003eFunding\\u003c/h2\\u003e\\u003cp\\u003eThis work was supported by the Scientific Research Project of Hubei Health Commission (Grant No. WJ2023M079) and Natural Science Foundation of Hubei (Grant No. 2024AFB707).\\u003c/p\\u003e\\u003ch2\\u003eAuthor Contribution\\u003c/h2\\u003e\\u003cp\\u003eAll of the authors have significantly contributed to this manuscript and are in agreement with its content. WZ and LWK conducted conceptualization and methodology. HPD was involved in both drafting of the original manuscript and generation of visual elements of the study. LJJ and YL were involving in the editing of the final manuscript as well as performing data analysis. All authors have read and agreed to the final version of the manuscript.\\u003c/p\\u003e\\u003ch2\\u003eAcknowledgement\\u003c/h2\\u003e\\u003cp\\u003eThe authors would like to express their sincere gratitude to the Scientific research project of Hubei Health Commission for the financial support.\\u003c/p\\u003e\\u003ch2\\u003eData Availability\\u003c/h2\\u003e\\u003cp\\u003eThe raw data supporting the findings of this study are available from the corresponding author upon reasonable request and without undue reservation.\\u003c/p\\u003e\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\u003cli\\u003e\\u003cspan\\u003eCampbell BCV, De Silva DA, Macleod MR, Coutts SB, Schwamm LH and Davis SM, et al. Ischaemic stroke. Nat Rev Dis Primers. 2019; 5(1): 70.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eSaini V, Guada L and Yavagal DR. Global epidemiology of stroke and access to acute ischemic stroke interventions. Neurology. 2021; 97(20 Suppl 2): S6-16.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eBonati LH, Gregson J, Dobson J, Mccabe DJH, Nederkoorn PJ and van der Worp HB, et al. Restenosis and risk of stroke after stenting or endarterectomy for symptomatic carotid stenosis in the international carotid stenting study (ICSS): secondary analysis of a randomised trial. Lancet Neurol. 2018; 17(7): 587\\u0026ndash;96.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eGolnari P, Nazari P, Hurley MC, Shaibani A, Ansari SA and Potts MB, et al. Continued dominance of carotid endarterectomy over stenting in the united states: volumes, outcomes, and complications from the national inpatient sample (1997\\u0026ndash;2015). World Neurosurg. 2022; 163: e238-52.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eGaba K, Ringleb PA and Halliday A. Asymptomatic carotid stenosis: intervention or best medical therapy? Curr Neurol Neurosci Rep. 2018; 18(11): 80.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eWhite CJ, Brott TG, Gray WA, Heck D, Jovin T and Lyden SP, et al. Carotid artery stenting: JACC state-of-the-art review. J Am Coll Cardiol. 2022; 80(2): 155\\u0026ndash;70.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eSwanson K, Wu E, Zhang A, Alizadeh AA and Zou J. From patterns to patients: advances in clinical machine learning for cancer diagnosis, prognosis, and treatment. Cell. 2023; 186(8): 1772\\u0026ndash;91.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eQin X, Yi S, Rong J, Lu H, Ji B and Zhang W, et al. Identification of anoikis-related genes classification patterns and immune infiltration characterization in ischemic stroke based on machine learning. Front Aging Neurosci. 2023; 15: 1142163.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eGene ontology consortium: going forward. Nucleic Acids Res. 2015; 43(Database issue): D1049-56.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eKanehisa M and Goto S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 2000; 28(1): 27\\u0026ndash;30.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eSubramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL and Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005; 102(43): 15545\\u0026ndash;50.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eBader GD and Hogue CWV. An automated method for finding molecular complexes in large protein interaction networks. BMC Bioinformatics. 2003; 4: 2.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eFurio-Tari P, Tarazona S, Gabaldon T, Enright AJ and Conesa A. Spongescan: a web for detecting microRNA binding elements in lncRNA sequences. Nucleic Acids Res. 2016; 44(W1): W176-80.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eZhou G, Soufan O, Ewald J, Hancock REW, Basu N and Xia J. NetworkAnalyst 3.0: a visual analytics platform for comprehensive gene expression profiling and meta-analysis. Nucleic Acids Res. 2019; 47(W1): W234-41.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eMaravic-Stojkovic V, Lausevic-Vuk LJ, Obradovic M, Jovanovic P, Tanaskovic S and Stojkovic B, et al. Copeptin level after carotid endarterectomy and perioperative stroke. Angiology. 2014; 65(2): 122\\u0026ndash;9.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eBarca C, Foray C, Hermann S, Herrlinger U, Remory I and Laoui D, et al. The colony stimulating factor-1 receptor (CSF-1r)-mediated regulation of microglia/macrophages as a target for neurological disorders (glioma, stroke). Front Immunol. 2021; 12: 787307.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eGutmann DH and Kettenmann H. Microglia/brain macrophages as central drivers of brain tumor pathobiology. Neuron. 2019; 104(3): 442\\u0026ndash;9.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eSpangenberg EE, Lee RJ, Najafi AR, Rice RA, Elmore MRP and Blurton-Jones M, et al. Eliminating microglia in alzheimer's mice prevents neuronal loss without modulating amyloid-beta pathology. Brain. 2016; 139(Pt 4): 1265\\u0026ndash;81.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eGinhoux F, Greter M, Leboeuf M, Nandi S, See P and Gokhan S, et al. Fate mapping analysis reveals that adult microglia derive from primitive macrophages. Science. 2010; 330(6005): 841\\u0026ndash;5.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eOtxoa-De-Amezaga A, Miro-Mur F, Pedragosa J, Gallizioli M, Justicia C and Gaja-Capdevila N, et al. Microglial cell loss after ischemic stroke favors brain neutrophil accumulation. Acta Neuropathol. 2019; 137(2): 321\\u0026ndash;41.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eWijeratne T, Menon R, Sales C, Karimi L and Crewther S. Carotid artery stenosis and inflammatory biomarkers: the role of inflammation-induced immunological responses affecting the vascular systems. Ann Transl Med. 2020; 8(19): 1276.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eLiu B, Su L, Loo SJ, Gao Y, Khin E and Kong X, et al. Matrix metallopeptidase 9 contributes to the beginning of plaque and is a potential biomarker for the early identification of atherosclerosis in asymptomatic patients with diabetes. Front Endocrinol (Lausanne). 2024; 15: 16.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eSomuncu MU, Pusuroglu H, Karakurt H, Bolat I, Karakurt ST and Demir AR, et al. The prognostic value of elevated matrix metalloproteinase-9 in patients undergoing primary percutaneous coronary intervention for ST-elevation myocardial infarction: a two-year prospective study. Rev Port Cardiol. 2020; 39(5): 267\\u0026ndash;76.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eCamare C, Pucelle M, Negre-Salvayre A and Salvayre R. Angiogenesis in the atherosclerotic plaque. Redox Biol. 2017; 12: 18\\u0026ndash;34.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eChen H, Valle G, Furlan S, Nani A, Gyorke S and Fill M, et al. Mechanism of calsequestrin regulation of single cardiac ryanodine receptor in normal and pathological conditions. J Gen Physiol. 2013; 142(2): 127\\u0026ndash;36.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eXiao Q, Hou R, Xie L, Niu M, Pan X and Zhu X. Macrophage metabolic reprogramming and atherosclerotic plaque microenvironment: fostering each other? Clin Transl Med. 2023; 13(5): e1257.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eDing Y, Yuan X, Wang Y and Yan J. CASQ2 alleviates lung cancer by inhibiting m2 tumor-associated macrophage polarization and JAK/STAT pathway. J Biochem Mol Toxicol. 2024; 38(8): e23801.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eShui L, Meng Y, Huang C, Qian Y and Liu J. Aminopeptidase n expression in the endometrium could affect endometrial receptivity. Biochem Biophys Res Commun. 2019; 514(2): 469\\u0026ndash;74.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eDevarakonda CV, Pereira FE, Smith JD, Shapiro LH and Ghosh M. CD13 deficiency leads to increased oxidative stress and larger atherosclerotic lesions. Atherosclerosis. 2019; 287: 70\\u0026ndash;80.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eDu H, Yang L, Zhang H, Zhang X and Shao H. LncRNA TUG1 silencing enhances proliferation and migration of ox-LDL-treated human umbilical vein endothelial cells and promotes atherosclerotic vascular injury repairing via the runx2/ANPEP axis. Int J Cardiol. 2021; 338: 204\\u0026ndash;14.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eMiao M, Wang X, Liu T, Li Y, Yu W and Yang T, et al. Targeting PPARs for therapy of atherosclerosis: a review. Int J Biol Macromol. 2023; 242(Pt 2): 125008.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eJin M, Fang J, Wang J, Shao X, Xu S and Liu P, et al. Regulation of toll-like receptor (TLR) signaling pathways in atherosclerosis: from mechanisms to targeted therapeutics. Acta Pharmacol Sin. 2023; 44(12): 2358\\u0026ndash;75.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eZhang Z, Yue P, Lu T, Wang Y, Wei Y and Wei X. Role of lysosomes in physiological activities, diseases, and therapy. J Hematol Oncol. 2021; 14(1): 79.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eChen P, Qin L, Li G, Wang Z, Dahlman JE and Malagon-Lopez J, et al. Endothelial TGF-beta signalling drives vascular inflammation and atherosclerosis. Nat Metab. 2019; 1(9): 912\\u0026ndash;26.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eLi L, Wang M, Ma Q, Ye J and Sun G. Role of glycolysis in the development of atherosclerosis. Am J Physiol Cell Physiol. 2022; 323(2): C617-29.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eChen H, Shi Z, Feng H, Wang R, Zhang Y and Xie J, et al. The relationship between dizziness and cervical artery stenosis. Neuroreport. 2015; 26(18): 1112\\u0026ndash;8.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eLiu H, Zhou J, Jiang W and Wang F. Analysis of the diagnostic and prognostic value of mir-9-5p in carotid artery stenosis. Bosn J Basic Med Sci. 2021; 21(6): 724\\u0026ndash;9.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eZhao G, Yu H, Ding L, Wang W, Wang H and Hu Y, et al. MicroRNA-27a-3p delivered by extracellular vesicles from glioblastoma cells induces m2 macrophage polarization via the EZH1/KDM3a/CTGF axis. Cell Death Discov. 2022; 8(1): 260.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eTeng L and Meng R. Long non-coding RNA MALAT1 promotes acute cerebral infarction through miRNAs-mediated hs-CRP regulation. J Mol Neurosci. 2019; 69(3): 494\\u0026ndash;504.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eZhu B, Liu W, Xu Q and Liu H. MicroRNA-486-5p functions as a diagnostic marker for carotid artery stenosis and prevents endothelial dysfunction through inhibiting inflammation and oxidative stress. Bioengineered. 2022; 13(4): 8667\\u0026ndash;75.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eVilla C and Stoccoro A. Epigenetic peripheral biomarkers for early diagnosis of alzheimer's disease. Genes (Basel). 2022; 13(8).\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eWang C, Zhao F, He Y, E Y and Li S. Long non-coding RNA RMST serves as a diagnostic biomarker in patients with carotid artery stenosis and predicts the occurrence of cerebral ischemic event: a retrospective study. Vascular. 2023; 31(5): 908\\u0026ndash;13.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eMa H and Dong A. Dysregulation of lncRNA SNHG1/mir-145 axis affects the biological function of human carotid artery smooth muscle cells as a mechanism of carotid artery restenosis. Exp Ther Med. 2021; 21(5): 423.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eWang F, Wang Y and Guo X. LncRNA THRIL functions as a marker for carotid artery stenosis and affects the biological function of human aortic endothelial cell. J Inflamm Res. 2023; 16: 2437\\u0026ndash;46.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eHao Z, Zhang Y, Li Y, Zhao J, Zhou Y and Qiu J, et al. The association between ideal cardiovascular health metrics and extracranial carotid artery stenosis in a northern chinese population: a cross-sectional study. Sci Rep. 2016; 6: 7.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eYan Y, Lu Y, Liu Y, Zhang J, Wang S and Dong J, et al. Identification of circular RNA hsa_circ_0034621 as a novel biomarker for carotid atherosclerosis and the potential function as a regulator of NLRP3 inflammasome. Atherosclerosis. 2024; 391: 10.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eZhang M, Zhu Y, Zhu J, Xie Y, Wu R and Zhong J, et al. Circ_0086296 induced atherosclerotic lesions via the IFIT1 / STAT1 feedback loop by sponging mir-576-3p. Cell Mol Biol Lett. 2022; 27(1): 26.\\u003c/span\\u003e\\u003c/li\\u003e\\u003cli\\u003e\\u003cspan\\u003eXie X, Huang C, Xu D, Liu Y, Hu M and Long J, et al. Elevation of hypertonicity - induced protein NFAT5 promotes apoptosis of human umbilical vein endothelial cells through the NF - κb pathway. Mol Med Rep. 2021; 23(3): 8.\\u003c/span\\u003e\\u003c/li\\u003e\\u003c/ol\\u003e\"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":false,\"hideJournal\":false,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":true,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"european-journal-of-medical-research\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":false,\"externalIdentity\":\"ejmr\",\"sideBox\":\"Learn more about [European Journal of Medical Research](http://eurjmedres.biomedcentral.com)\",\"snPcode\":\"40001\",\"submissionUrl\":\"https://submission.nature.com/new-submission/40001/3\",\"title\":\"European Journal of Medical Research\",\"twitterHandle\":\"@BioMedCentral\",\"acdcEnabled\":true,\"dfaEnabled\":true,\"editorialSystem\":\"em\",\"reportingPortfolio\":\"BMC/SO AJ\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":true},\"keywords\":\"Atherosclerotic carotid artery stenosis, Pathogenic markers, Immune cell correlation, Machine learning\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-7112702/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-7112702/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003e\\u003cstrong\\u003eObjective:\\u003c/strong\\u003e A comprehensive bioinformatics analysis was conducted to identify key genes and regulatory networks associated with atherosclerotic carotid artery stenosis (ACAS).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eMethods: \\u003c/strong\\u003eFour datasets, including GSE43292, GSE100927, GSE28829, and GSE198600, were integrated to form the training set, with the GSE163154 dataset serving as the validation set. Subsequently, differential expression and functional enrichment analysis were performed on the training set. Additionally, key pathogenic genes were identified using the protein-protein interaction networks, molecular complex detection technique, and three machine learning (ML) algorithms. These identified genes were validated through inter-group differences and receiver operating characteristic (ROC) curve analyses. Immune-related functions and immune cell correlations were analyzed and verified using ACAS plaque tissue samples.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eResults: \\u003c/strong\\u003eFollowing the analysis, a total of 33 downregulated and 52 upregulated genes were identified. Furthermore, enrichment analysis of gene sets demonstrated that the highly expressed group was involved in cellular receptor signaling, leishmaniasis infection, lysosome, PPAR-signaling, and Toll-like receptor pathways. In contrast, the low-expressed group was involved in mechanisms involving dilated cardiomyopathy, pyruvate metabolism, hypertrophic cardiomyopathy, spliceosome, and TGF-β signaling pathways. Notably, ANPEP, CSF1R, MMP9, and CASQ2 were found to differ significantly between groups. Correlation analysis revealed positive associations between MMP9 expression and neutrophil infiltration, CASQ2 expression and M2 macrophage abundance, and CSF1R expression and M1 macrophage levels.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eConclusion: \\u003c/strong\\u003eConsequently, these genes may serve as potential biomarkers and therapeutic targets in the diagnosis and treatment of ACAS.\\u003c/p\\u003e\",\"manuscriptTitle\":\"Identification of key genes and regulatory networks associated with atherosclerotic carotid artery stenosis through comprehensive bioinformatics analysis and machine learning\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2025-08-11 14:41:28\",\"doi\":\"10.21203/rs.3.rs-7112702/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0},{\"type\":\"decision\",\"content\":\"Revision requested\",\"date\":\"2025-09-17T18:34:16+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"editorInvitedReview\",\"content\":\"\",\"date\":\"2025-08-26T04:47:09+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"editorInvitedReview\",\"content\":\"\",\"date\":\"2025-08-23T11:56:35+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"editorInvitedReview\",\"content\":\"\",\"date\":\"2025-08-13T18:47:29+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewerAgreed\",\"content\":\"145983098114690182186250791040280634815\",\"date\":\"2025-08-08T12:16:36+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewerAgreed\",\"content\":\"28625242907170448776010913368445526033\",\"date\":\"2025-08-07T03:49:26+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewerAgreed\",\"content\":\"253466424409666898087135777852860345723\",\"date\":\"2025-08-06T18:55:35+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewersInvited\",\"content\":\"\",\"date\":\"2025-08-06T12:12:25+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"editorAssigned\",\"content\":\"\",\"date\":\"2025-07-16T15:30:41+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"checksComplete\",\"content\":\"\",\"date\":\"2025-07-15T05:32:03+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"submitted\",\"content\":\"European Journal of Medical Research\",\"date\":\"2025-07-13T10:22:56+00:00\",\"index\":\"\",\"fulltext\":\"\"}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"european-journal-of-medical-research\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":false,\"externalIdentity\":\"ejmr\",\"sideBox\":\"Learn more about [European Journal of Medical Research](http://eurjmedres.biomedcentral.com)\",\"snPcode\":\"40001\",\"submissionUrl\":\"https://submission.nature.com/new-submission/40001/3\",\"title\":\"European Journal of Medical Research\",\"twitterHandle\":\"@BioMedCentral\",\"acdcEnabled\":true,\"dfaEnabled\":true,\"editorialSystem\":\"em\",\"reportingPortfolio\":\"BMC/SO AJ\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"14b45d71-ccb8-4ac4-97ba-2f1d0d7bc34a\",\"owner\":[],\"postedDate\":\"August 11th, 2025\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"published-in-journal\",\"subjectAreas\":[],\"tags\":[],\"updatedAt\":\"2025-11-10T16:04:51+00:00\",\"versionOfRecord\":{\"articleIdentity\":\"rs-7112702\",\"link\":\"https://doi.org/10.1186/s40001-025-03330-8\",\"journal\":{\"identity\":\"european-journal-of-medical-research\",\"isVorOnly\":false,\"title\":\"European Journal of Medical Research\"},\"publishedOn\":\"2025-11-04 15:57:38\",\"publishedOnDateReadable\":\"November 4th, 2025\"},\"versionCreatedAt\":\"2025-08-11 14:41:28\",\"video\":\"\",\"vorDoi\":\"10.1186/s40001-025-03330-8\",\"vorDoiUrl\":\"https://doi.org/10.1186/s40001-025-03330-8\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-7112702\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-7112702\",\"identity\":\"rs-7112702\",\"version\":[\"v1\"]},\"buildId\":\"8U1c8b4HqxoKbykW_rLl7\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}