Incorporating in-source fragment information improves metabolite identification accuracy in untargeted LC-MS datasets.

In-source fragmentation occurs as a byproduct of electrospray ionization. We find that ions produced as a result of in-source fragmentation often match fragment ions produced during MS/MS fragmentation, and we take advantage of this phenomenon in a novel algorithm to analyze LC-MS metabolomics data sets. Our approach organizes coeluting MS1 features into a single peak group and then identifies in-source fragments among coeluting features using MS/MS spectral libraries. We tested our approach using previously published data of verified metabolites and compared the results to features detected by other mainstream metabolomics tools. Our results indicate that considering in -source fragment information as a part of the identification process increases the annotation quality, allowing us to leverage MS/MS data in spectrum libraries even if MS/MS scans were not collected.