Scholarlink

​Background <p>Ischemic stroke (IS) and aging share similar pathophysiological features, including vascular dysfunction, inflammatory responses, and oxidative stress. However, the underlying molecular mechanisms remain unclear. This study aimed to identify key shared drive genes between IS and aging through integrated multi-omics analysis and machine learning approaches, and to explore their diagnostic and therapeutic potential.</p> ​Methods <p>Based on the GSE22255 and GSE58294 datasets, differentially expressed genes (DEGs) associated with IS were screened using differential expression analysis and weighted gene co-expression network analysis (WGCNA). The overlapping genes between IS-related DEGs and aging-related genes (ARGs) were identified as aging-related DEGs (ARDEGs). Further refinement of core genes was performed through Gene Ontology (GO)/Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis, protein-protein interaction (PPI) network construction, and machine learning algorithms (LASSO regression and support vector machine [SVM]). Immune cell infiltration was evaluated using CIBERSORT, and gene expression patterns were validated via the DISCO single-cell database. Finally, molecular docking and drug database screening were employed to predict potential therapeutic targets.</p> ​Results <p>A total of 279 IS-related DEGs were identified, among which 29 showed significant overlap with ARGs (ARDEGs). WGCNA revealed that the MEcyan module exhibited a strong negative correlation with IS phenotypes (r = -0.63). Machine learning algorithms identified JUP, UQCRC1, and MRPL41 as core potential diagnostic biomarkers, with area under the curve (AUC) values all exceeding 0.7. Immune infiltration analysis demonstrated a significant increase in M1 macrophages and neutrophils, along with a reduction in CD8+ T cells in IS patients. Single-cell data confirmed the specific expression of these core genes in neurons and immune cells. Molecular docking suggested that the herbicide atrazine may target these genes (binding energy &lt; -5.7 kcal/mol), while hsa-miR-30c-5p was predicted to regulate JUP and UQCRC1.</p> ​Conclusion <p>This study elucidates key shared genes and immune microenvironment features between IS and aging, proposing JUP, UQCRC1, and MRPL41 as potential diagnostic biomarkers. Furthermore, atrazine was identified in silicoas a potential interacting molecule, suggesting the druggability of these target proteins​ and providing a starting point for future drug discovery efforts in IS.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

​Aging-related gene signatures as potential biomarkers in ischemic stroke: an integrated bioinformatics and machine learning study

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月17日

查看全文
  • Peilu Wang,
  • Yan Wang,
  • Dongliang Wang,
  • Yingxuan Li,
  • Zhenyu Yin,
  • Fuqing Zhang

摘要

​Background

Ischemic stroke (IS) and aging share similar pathophysiological features, including vascular dysfunction, inflammatory responses, and oxidative stress. However, the underlying molecular mechanisms remain unclear. This study aimed to identify key shared drive genes between IS and aging through integrated multi-omics analysis and machine learning approaches, and to explore their diagnostic and therapeutic potential.

​Methods

Based on the GSE22255 and GSE58294 datasets, differentially expressed genes (DEGs) associated with IS were screened using differential expression analysis and weighted gene co-expression network analysis (WGCNA). The overlapping genes between IS-related DEGs and aging-related genes (ARGs) were identified as aging-related DEGs (ARDEGs). Further refinement of core genes was performed through Gene Ontology (GO)/Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis, protein-protein interaction (PPI) network construction, and machine learning algorithms (LASSO regression and support vector machine [SVM]). Immune cell infiltration was evaluated using CIBERSORT, and gene expression patterns were validated via the DISCO single-cell database. Finally, molecular docking and drug database screening were employed to predict potential therapeutic targets.

​Results

A total of 279 IS-related DEGs were identified, among which 29 showed significant overlap with ARGs (ARDEGs). WGCNA revealed that the MEcyan module exhibited a strong negative correlation with IS phenotypes (r = -0.63). Machine learning algorithms identified JUP, UQCRC1, and MRPL41 as core potential diagnostic biomarkers, with area under the curve (AUC) values all exceeding 0.7. Immune infiltration analysis demonstrated a significant increase in M1 macrophages and neutrophils, along with a reduction in CD8+ T cells in IS patients. Single-cell data confirmed the specific expression of these core genes in neurons and immune cells. Molecular docking suggested that the herbicide atrazine may target these genes (binding energy < -5.7 kcal/mol), while hsa-miR-30c-5p was predicted to regulate JUP and UQCRC1.

​Conclusion

This study elucidates key shared genes and immune microenvironment features between IS and aging, proposing JUP, UQCRC1, and MRPL41 as potential diagnostic biomarkers. Furthermore, atrazine was identified in silicoas a potential interacting molecule, suggesting the druggability of these target proteins​ and providing a starting point for future drug discovery efforts in IS.