MS2Query: reliable and scalable MS2 mass spectra-based analogue search

Abstract

The authors develop a machine learning approach to find structurally related chemicals in mass spectral libraries. Their method boosts the annotation rate and aids in assessing novelty in metabolomics datasets.