Abstract
Liquid chromatography-mass spectrometry (LC-MS) is a key technology in metabolomics, enabling high-throughput detection of small molecules across diverse biological samples. However, raw LC-MS data are complex, requiring careful preprocessing to ensure accurate and reproducible feature detection. This chapter introduces a step-by-step protocol for LC-MS data preprocessing using the open-source xcms package in R. Designed for users ranging from beginners to experienced analysts, the chapter outlines critical stages including data inspection, peak detection, retention time alignment, correspondence, and result export. Special attention is given to parameter optimization and diagnostic visualization to guide users in making informed decisions tailored to their specific datasets. We demonstrate the approach on human serum samples, using real-life example compounds such as proline to showcase retention time alignment and peak correspondence. By combining practical code snippets with conceptual insights, this chapter empowers researchers to harness xcms for robust, reproducible LC-MS workflows in untargeted metabolomics. Whether you are setting up your first analysis or refining an established pipeline, this chapter serves as both a tutorial and a reference for high-quality LC-MS data processing.