IsoPairFinder: A Computational Workflow for Prioritizing LC–MS Feature Pairs Representing Candidate Microbial Pathway Intermediates from Stable Isotope Tracing Metabolomics

Abstract Functional annotation of microbial genes lags behind the pace of genome sequencing, leaving major gaps in our understanding of microbial metabolism. Combining genetic perturbation with untargeted stable isotope tracing metabolomics provides a powerful strategy for identifying pathway intermediates and elucidating microbial metabolic pathways. To streamline this workflow, we developed IsoPairFinder, a computational tool that prioritizes pathway-intermediate feature pairs by analyzing paired unlabeled (12C) and 13C-labeled metabolomics datasets from genetically perturbed microbes. By consolidating redundant liquid chromatography–mass spectrometry (LC–MS) features and pairing features across unlabeled and labeled samples, IsoPairFinder reduces candidate lists to plausible isotope-paired intermediates for downstream validation. We demonstrate the utility of IsoPairFinder in a case study of uric acid catabolism in Clostridium sporogenes, where it filters redundant features and prioritizes LC–MS feature pairs representing gene-associated pathway intermediates. IsoPairFinder is available as an open-source R package and as a web-based workflow within GNPS2. By integrating differential analysis and isotope-pair detection into a single workflow, IsoPairFinder enables microbiologists to connect metabolomic features with microbial gene function more efficiently, thereby accelerating microbial pathway discovery.