Department of Chemistry, Institutes of Biomedical Sciences, Zhongshan Hospital, Fudan University, Shanghai 200433, China.
School of Materials Science and Chemical Engineering, Ningbo University, Ningbo 315211, China.
Anal Chem. 2024 Sep 10;96(36):14688-14696. doi: 10.1021/acs.analchem.4c03681. Epub 2024 Aug 29.
Metabolomics analysis based on body fluids, combined with high-throughput laser desorption and ionization mass spectrometry (LDI-MS), holds great potential and promising prospects for disease diagnosis and screening. On the other hand, chronic obstructive pulmonary disease (COPD) currently lacks innovative and powerful diagnostic and screening methods. In this work, CoFeNMOF-D, a metal-organic framework (MOF)-derived metal oxide nanomaterial, was synthesized and utilized as a matrix to assist LDI-MS for extracting serum metabolic fingerprints of COPD patients and healthy controls (HC). Through machine learning algorithms, successful discrimination between the COPD and HC was achieved. Furthermore, four potential biomarkers significantly downregulated in COPD were screened out. The disease diagnostic models based on the biomarkers demonstrated excellent diagnostic performance across different algorithms, with area under the curve (AUC) values reaching 0.931 and 0.978 in the training and validation sets, respectively. Finally, the potential metabolic pathways and disease mechanisms associated with the identified markers were explored. This work advances the application of LDI-based molecular diagnostics in clinical settings.
基于体液的代谢组学分析结合高通量激光解吸电离质谱（LDI-MS），为疾病的诊断和筛查提供了巨大的潜力和广阔的前景。另一方面，慢性阻塞性肺疾病（COPD）目前缺乏创新和强大的诊断和筛查方法。在这项工作中，合成了一种金属-有机骨架（MOF）衍生的金属氧化物纳米材料 CoFeNMOF-D，并将其用作基质，辅助 LDI-MS 提取 COPD 患者和健康对照（HC）的血清代谢指纹图谱。通过机器学习算法，成功地区分了 COPD 和 HC。此外，筛选出了四个在 COPD 中显著下调的潜在生物标志物。基于生物标志物的疾病诊断模型在不同的算法中表现出了出色的诊断性能，在训练集和验证集中的 AUC 值分别达到 0.931 和 0.978。最后，探讨了与鉴定标志物相关的潜在代谢途径和疾病机制。这项工作推进了基于 LDI 的分子诊断在临床中的应用。
