Deep Learning-Enabled\nRapid Metabolic Decoding of\nSmall Extracellular Vesicles via Dual-Use Mass Spectroscopy Chip Array

The increasing focus of small extracellular vesicles\n(sEVs) in\nliquid biopsy has created a significant demand for streamlined improvements\nin sEV isolation methods, efficient collection of high-quality sEV\ndata, and powerful rapid analysis of large data sets. Herein, we develop\na high-throughput dual-use mass spectroscopic chip array (DUMSCA)\nfor the rapid isolation and detection of plasma sEVs. The DUMSCA realizes\nmore than a 50% increase in speed compared to traditional method and\nconfirms proficiency in robust storage, reuse, high-efficiency desorption/ionization,\nand metabolite quantification. With the collected metabolic data matrix\nof sEVs, a deep learning model achieves high-performance diagnosis\nof Crohn’s disease. Furthermore, discovered biomarkers by feature\nsparsification and tandem mass spectrometry experiments also exhibited\nremarkable performance in diagnosis. This work demonstrates the rapidity\nand validity of DUMSCA for disease diagnosis, enabling the diagnosis\nof diseases without the necessity for prior knowledge and providing\na high-throughput technology for sEV-based liquid biopsy that will\nempower its vigorous development.