Scholarlink

<p>Characterizing the isomeric diversity of molecular formulas (MFs) in dissolved organic matter (DOM) is essential for advancing research across environmental and biomedical sciences. Ultra-high-resolution mass spectrometry (UHR MS), particularly when coupled with high-performance liquid chromatography (LC-UHR MS), can resolve isomeric diversity chromatographically. However, its broad application is limited by high operational costs, the requirement for expert handling, and complex data interpretation. Here, we present GUIDE (Graph-based Understanding of Isomeric Diversity), a predictive framework that infers isomeric diversity directly from direct infusion UHRMS (DI-UHR MS) data, effectively bridging the gap between LC-UHR MS and DI-UHR MS. The framework incorporates a self-supervised graph learning module to learn MF representations from intrinsic molecular features and spatial topological relationships, followed by a deep neural networks for isomeric diversity prediction. Our approach achieves high accuracy in predicting chromatographic isomeric diversity of natural DOM, with consistent performance across multiple DI-UHR MS platforms, including DI-FT-ICR MS and DI-Orbitrap MS. This approach enables the resolution of MFs at the isomeric level, offering a refined molecular perspective of DOM composition and opening new avenues for research in biogeochemistry, environmental science, and analytical chemistry.</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

Graph-based understanding of isomeric diversity in complex dissolved organic matter

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月27日

查看全文
  • Tongcun Liu,
  • Yuanbi Yi,
  • Chen Zhao,
  • Zekun Zhang,
  • Julian Merder,
  • Andrew J. Tanentzap,
  • Zhenwei Yan,
  • Fan Zhang,
  • Hailin Feng,
  • Can Yang,
  • Ding He

摘要

Characterizing the isomeric diversity of molecular formulas (MFs) in dissolved organic matter (DOM) is essential for advancing research across environmental and biomedical sciences. Ultra-high-resolution mass spectrometry (UHR MS), particularly when coupled with high-performance liquid chromatography (LC-UHR MS), can resolve isomeric diversity chromatographically. However, its broad application is limited by high operational costs, the requirement for expert handling, and complex data interpretation. Here, we present GUIDE (Graph-based Understanding of Isomeric Diversity), a predictive framework that infers isomeric diversity directly from direct infusion UHRMS (DI-UHR MS) data, effectively bridging the gap between LC-UHR MS and DI-UHR MS. The framework incorporates a self-supervised graph learning module to learn MF representations from intrinsic molecular features and spatial topological relationships, followed by a deep neural networks for isomeric diversity prediction. Our approach achieves high accuracy in predicting chromatographic isomeric diversity of natural DOM, with consistent performance across multiple DI-UHR MS platforms, including DI-FT-ICR MS and DI-Orbitrap MS. This approach enables the resolution of MFs at the isomeric level, offering a refined molecular perspective of DOM composition and opening new avenues for research in biogeochemistry, environmental science, and analytical chemistry.