Multidimensional Mass Spectral Similarity Algorithm: Discriminate Disaccharide and Flavonoid Isomers Coupled with...

Structural isomers are critical analytes in the biological and chemical arenas. Despite the ability of tandem mass spectrometry to provide fragment ion information, their high structural similarity impedes confident identification. To address this, we developed a novel method leveraging energy-resolved mass spectrometry (ER-MS) of fragment ions generated by electron activation dissociation (EAD). EAD initiated rapid radical chain dissociation via electron excitation and removal mechanisms, delivering superior isomer discrimination compared to conventional collision-induced dissociation (CID). Subsequent energy-resolved analysis further enhanced the distinction by integrating these dissociation mechanisms. Our strategy employed a cosine-based multidimensional spectral similarity algorithm to visualize and quantify subtle spectral differences across multiple energies. This method successfully distinguished many types of isomers, such as linkage, composition, and conformation isomers in disaccharides and flavonoid glycosides and achieved 93.8% top-1 identification accuracy against an in-house library. When applied to pomelo peel and commercial beverages for key metabolite characterization, it provided 44.4-50.0% top-1 annotation accuracy across all detected interest features. These results demonstrate that the multidimensional similarity algorithm that combines EAD and ER-MS significantly advances the depth and accuracy of compound annotation.