A Practical Guide to Dereplication in Natural Products Metabolomics

Abstract

Liquid chromatography-tandem mass spectrometry (LC−MS/MS)-based metabolomics has become a central approach for characterizing the chemical diversity of biological systems, particularly in natural product discovery. However, the scale and complexity of untargeted LC−MS/MS datasets present a persistent challenge: only a small fraction of detected features can be confidently assigned to known structures, leaving much of the metabolome unexplored. Dereplication and annotation tools have rapidly expanded to address this bottleneck, including molecular networking, spectral library matching, substructure discovery, chemical class prediction, in silico structure prediction, and repository-scale spectral searching. For beginners, these growing computational tools can be difficult to navigate since individual tools differ in input requirements, confidence levels, and optimal use cases. This review provides a practical overview of major annotation strategies in LC−MS/MS-based metabolomics and emphasizes how complementary tools can be integrated into coherent workflows. We first introduce core approaches for feature organization and dereplication, then discuss advanced strategies for substructure analysis, chemical classification, de novo annotation, and genome-informed discovery. Finally, representative applications from studies on different biological systems illustrate how integrative workflows enable prioritization, structural characterization, and discovery of previously unknown chemistry and biochemical reactions.