Answers are grounded in this paper and cite their sources.
Metabolomics is recognized as a crucial scientific domain, promising to advance our understanding of cell biology, physiology and medicine. Tandem mass spectrometry (MS/MS) has strong potential to elucidate the metabolites in complex biological samples and to become a standard tool complementing established techniques. However, despite its potential for answering many key questions, a major challenge in the use of tandem mass spectrometry for characterizing metabolites lies in a lack of computational tools for accurate annotation and structure identification allowing us to turn complex data into molecular knowledge. Chemo‐informatics and related machine‐learning in silico fragmentation tools have already been established and used for different classes of metabolites. For the classes of metabolites where existing chemo‐informatics approaches produce insufficiently accurate predictions a supervised machine learning based strategy can be used to predict possible molecular structures from “unassigned” experimental tandem MS data. Here, we propose a new innovative in silico approach employing quantum mechanical (QM) methods in order to predict ion formation and subsequent fragmentation patterns of arbitrary small molecules and validate putative annotations of tandem mass spectrometry (MS) data. The focus is on the evaluation of a new conceptual density functional theory (CDFT) nuclear reactivity descriptor of the nuclear Fukui function type, that characterizes the forces that the atomic nuclei experience due to proton attachment and captures the onset of the change in the nuclear positions induced by it. A series of test compounds for which high quality experimental data exist and that were investigated before in a more approximate theoretical framework have been examined. The output of these calculations provides a list of the most probable molecular structures predicted to match the experimental tandem MS spectrum (“de novo metabolite identification”).
No takes yet. Share an insight, caveat, or question.
Cauët et al. (2020) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: