| Literature DB >> 30103269 |
Jördis-Ann Schüler1,2, Steffen Neumann3,4, Matthias Müller-Hannemann1, Wolfgang Brandt2.
Abstract
Identification and structural determination of small molecules by mass spectrometry is an important step in chemistry and biochemistry. However, the chemically realistic annotation of a fragment ion spectrum can be a difficult challenge. We developed ChemFrag, for the detection of fragmentation pathways and the annotation of fragment ions with chemically reasonable structures. ChemFrag combines a quantum chemical with a rule-based approach. For different doping substances as test instances, ChemFrag correctly annotates fragment ions. In most cases, the predicted fragments are chemically more realistic than those from purely combinatorial approaches, or approaches based on machine learning. The annotation generated by ChemFrag often coincides with spectra that have been manually annotated by experts. This is a major advance in peak annotation and allows a more precise automatic interpretation of mass spectra.Keywords: doping substances; fragment ion annotation; rule-based fragmentation; semi-empirical quantum mechanics
Year: 2018 PMID: 30103269 DOI: 10.1002/jms.4278
Source DB: PubMed Journal: J Mass Spectrom ISSN: 1076-5174 Impact factor: 1.982