Interpretable Machine Learning-based Automated HPLC/MS2 Platform using Ion-Molecule Reactions for the Identification of Functionalities in Analytes

Abstract

Identification of unknown compounds in complex mixtures is a time-consuming and challenging problem in several areas of chemistry. High-performance liquid chromatography (HPLC) coupled to tandem mass spectrometry (MS2) based on collision-activated dissociation (CAD) is a standard approach used to identify unknown compounds in complex mixtures. However, CAD often produces similar fragmentation patterns for isomeric or related ionized analytes, which makes it difficult to differentiate between similar ions. MS/MS methods based on diagnostic gas-phase ion-molecule reactions provide a powerful, predictable, and reliable alternative for the differentiation of isomeric or similar ions via the identification of their specific functional groups. However, the interpretation of the experimental results, the selection of appropriate neutral reagents for new analytes, and the optimization of the conditions for reagent introduction is a manual, time-consuming and challenging process. We have developed a chemical graph-based interpretable machine learning approach that enables automated identification of functionalities in previously unknown protonated analytes, which facilitates the differentiation of isomeric or otherwise similar compounds. Furthermore, this approach significantly advances prior methods used to study ion–molecule reactions by enabling, for the first time, automated selection of neutral reagents for previously unstudied analytes and algorithmic optimization of reagent-dependent pulsing-in and pumping-out times for reagents introduced into the mass spectrometer via pulsed valves. This study establishes a foundation for fully automated HPLC/MS2 platforms, enabling the differentiation of similar unknown compounds in complex mixtures with broad applications across chemical sciences.