Scholarlink

<p>Mitochondrial DNA methylation (mtDNA-M) is a crucial epigenetic factor in ischemic stroke (IS); however, its specific role in the disease remains poorly characterized. This study aimed to explore MTDM-related key genes and therapeutic targets in IS through bioinformatics, machine learning, and experimental validation. Differentially expressed genes were identified using the GSE16561 dataset, and MTDM-related molecular subtypes of IS were classified via consensus clustering based on MTDM-related genes. We identified 52 candidate genes through the intersection of DEGs across subtypes and cases, from which seven key genes were prioritized through three machine learning algorithms (LASSO, SVM-RFE, and Boruta) to identify the most robust candidates. Among them, CD6, EVL, HIST2H2AA3, and SORL1 showed consistent robust discriminatory power (AUC &gt; 0.7) and were validated using BP neural networks and RT-qPCR in clinical samples. Functional enrichment analysis revealed their involvement in immune-related pathways, particularly in CD8 + T cell infiltration and primary immunodeficiency. Gene co-expression and regulatory network mapping suggested key miRNA (e.g., hsa-miR-26a-5p) and lncRNA (e.g., KCNQ1OT1) interactions, proposing layered transcriptional control. Our findings identify four immune-related genes as potential key genes and therapeutic targets for precision medicine in IS, advancing our understanding of the disease’s epigenetic regulation.</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

Integrated bulk and single-cell RNA-seq data identifies and validates key genes of mitochondrial-nuclear epigenetic axis in ischemic stroke

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月29日

查看全文
  • Shuangshuang Wang,
  • Yunke Zhang,
  • Bo Feng,
  • Guofang Yang,
  • Tingting Wang

摘要

Mitochondrial DNA methylation (mtDNA-M) is a crucial epigenetic factor in ischemic stroke (IS); however, its specific role in the disease remains poorly characterized. This study aimed to explore MTDM-related key genes and therapeutic targets in IS through bioinformatics, machine learning, and experimental validation. Differentially expressed genes were identified using the GSE16561 dataset, and MTDM-related molecular subtypes of IS were classified via consensus clustering based on MTDM-related genes. We identified 52 candidate genes through the intersection of DEGs across subtypes and cases, from which seven key genes were prioritized through three machine learning algorithms (LASSO, SVM-RFE, and Boruta) to identify the most robust candidates. Among them, CD6, EVL, HIST2H2AA3, and SORL1 showed consistent robust discriminatory power (AUC > 0.7) and were validated using BP neural networks and RT-qPCR in clinical samples. Functional enrichment analysis revealed their involvement in immune-related pathways, particularly in CD8 + T cell infiltration and primary immunodeficiency. Gene co-expression and regulatory network mapping suggested key miRNA (e.g., hsa-miR-26a-5p) and lncRNA (e.g., KCNQ1OT1) interactions, proposing layered transcriptional control. Our findings identify four immune-related genes as potential key genes and therapeutic targets for precision medicine in IS, advancing our understanding of the disease’s epigenetic regulation.