A compounds annotation strategy using targeted molecular networking for offline two-dimensional liquid chromatography-mass spectrometry analysis: Yupingfeng as a case study

Abstract

Component overlapping and long-time consumption hinder the data processing of offline two-dimensional liquid chromatography mass spectrometry (offline 2D-LC MS) system. Although molecular networking has been commonly employed in data processing of liquid chromatography mass spectrometry (LC-MS), its application in offline 2D-LC MS is challenged by voluminous and redundant data. In light of this, for the first time, a data deduplication and visualization strategy combining hand-in-hand alignment with targeted molecular networking (TMN) for compounds annotation of offline 2D-LC MS data was developed and applied to the chemical profile of Yupingfeng (YPF), a classical traditional Chinese medicine (TCM) prescription, as a case study. Firstly, an offline 2D-LC MS system was constructed for the separation and data acquisition of YPF extract. Then the data of 12 fractions derived from YPF were deconvoluted and aligned as a whole data file by hand-in-hand alignment, resulting in a 49.2% reduction in component overlapping (from 17951 to 9112 ions) and an improvement in the MS2 spectrum quality of precursor ions. Subsequently, the MS2-similarity adjacency matrix of focused parent ions was computed by a self-building Python script, which realized the construction of an innovative TMN. Interestingly, the TMN was found to be able to efficiently distinguish and visualize the co-elution, in-source fragmentations and multi-type adduct ions in a clustering network. Consequently, a total of 497 compounds were successfully identified depending on only seven TMN analysis guided by product ions filtering (PIF) and neutral loss filtering (NLF) for the targeted compounds in YPF. This integrated strategy improved the efficiency of targeted compound discovery in offline 2D-LC MS data, also shown a huge scalability in accurate compound annotation of complex samples. In conclusion, our study developed available concepts and tools while providing a research paradigm for efficient and rapid compound annotation in complex samples such as TCM prescriptions, with YPF as an example.