Scholarlink

<p>Major depressive disorder (MDD) is a leading cause of disability worldwide, yet its neural basis and biomarkers remain elusive. In this study, we combined neuroimaging, transcriptomics, proteomics, and machine learning to bridge large-scale brain network dysfunction with peripheral molecular signatures. Resting-state functional magnetic resonance imaging (fMRI) from 544 MDD patients and 569 healthy controls (HC) revealed altered network topology involving thalamic, fronto-striatal, and parietal regions. Connectome–transcriptomic association analysis of these network alterations, in conjunction with differential gene expression analysis, identified 24 candidate genes, with FKBP5, PTX3, and APCDD1 prioritized by machine learning. Plasma proteomics further validated altered protein expression of the three prioritized genes in two independent cohorts (adults: 162 MDD patients and 103 HC; adolescents: 71 MDD patients and 39 HC). Protein-based diagnostic models achieved accuracies of 69.6% in adults and 83.1% in adolescents, while protein levels were significantly higher in adolescents with MDD than in HC (all P &lt; 0.05). These findings suggest that PTX3, APCDD1, and FKBP5 may serve as candidate peripheral biomarkers associated with MDD-related brain network alterations.</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

From brain networks to peripheral signatures: candidate biomarkers for major depressive disorder

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月27日

查看全文
  • Yuan Liu,
  • Meijuan Li,
  • Bo Yan,
  • Ying Gao,
  • Yifan Jing,
  • Chengfeng Chen,
  • Xue Tian,
  • Yan Zhou,
  • Mengxin Xie,
  • Changlin Zhang,
  • Bin Zhang,
  • Jie Li

摘要

Major depressive disorder (MDD) is a leading cause of disability worldwide, yet its neural basis and biomarkers remain elusive. In this study, we combined neuroimaging, transcriptomics, proteomics, and machine learning to bridge large-scale brain network dysfunction with peripheral molecular signatures. Resting-state functional magnetic resonance imaging (fMRI) from 544 MDD patients and 569 healthy controls (HC) revealed altered network topology involving thalamic, fronto-striatal, and parietal regions. Connectome–transcriptomic association analysis of these network alterations, in conjunction with differential gene expression analysis, identified 24 candidate genes, with FKBP5, PTX3, and APCDD1 prioritized by machine learning. Plasma proteomics further validated altered protein expression of the three prioritized genes in two independent cohorts (adults: 162 MDD patients and 103 HC; adolescents: 71 MDD patients and 39 HC). Protein-based diagnostic models achieved accuracies of 69.6% in adults and 83.1% in adolescents, while protein levels were significantly higher in adolescents with MDD than in HC (all P < 0.05). These findings suggest that PTX3, APCDD1, and FKBP5 may serve as candidate peripheral biomarkers associated with MDD-related brain network alterations.