Mass Spectrometry‐Based Chromatographic and Computational Workflows for Biomarker and Therapeutic Target Discovery: A Comprehensive Review

The identification of biomarkers and therapeutic targets is essential to the success of precision medicine. However, to reliably identify them, one must employ a combination of analytical and computational methods. This article describes the importance of high-quality study design, strict adherence to preanalytical and quality assurance practices and high levels of quality assurance as well as the performance characteristics of multiple types of chromatography/mass spectrometry to achieve the best possible results for sensitive, selective biomarker profiling in various biological fluids such as plasma, serum, urine, cerebrospinal fluid (CSF), and tissues. In addition, it explores the key workflows used in preparing biological samples (i.e., solid-phase extraction/specialised physical or chemical extractions/derivatisation), software/data processing pipelines for peak analysis (i.e., XCMS/MZmine/OpenMS/MS-DIAL) and processes for the identification of compounds by combining spectral libraries (e.g., Human Metabolome Database [HMDB], National Institute of Standards and Technology [NIST] and FiehnLib) with in silico tools (e.g., SIRIUS, CSI: FingerID, CANOPUS). Finally, it discusses several ways in which artificial intelligence/machine learning can be applied to the field of mass spectrometry for peak detection and provides a comprehensive review of the requirements for regulatory-grade validation and the workflow for targeted/multiplexed/quantitative liquid chromatography/mass spectrometry (LC/MS) and gas chromatography/mass spectrometry (GC-MS) using stable isotope-labelled internal standards.