Modified two-dose difference with stable isotope tracing workflow using mass shift defect filtering substantially reduces false positives in untargeted nifedipine metabolite profiling.

BACKGROUND: Untargeted drug-metabolite identification continues to be challenged by the tradeoff between comprehensive coverage and acceptable false-positive rates. Our previously developed two-dose-difference platform combined with stable isotope tracing (SIT) achieved highly comprehensive and accurate metabolite detection. However, this approach still produced a large false-positive rate. Thus, it is critical to develop strategies that reduce the false-positive burden.

RESULTS: We benchmarked four data-processing workflows-original two-dose-difference + SIT, modified two-dose-difference + SIT (with a mass shift defect filter), dose-response + SIT, and mass defect filtering + SIT-using nifedipine (NIF) as a probe. UPLC-MS data were acquired under three incubation setups: coincubation of NIF with isotope-labeled analog (D4-NIF), separate incubations, and postreaction mixing of supernatants. Targeted MS/MS validation confirmed 65 putative NIF metabolites, including three previously reported. Among the four data-processing workflows, the original and modified two-dose-difference + SIT workflows proved to be the most comprehensive for NIF metabolite identification. Compared with the original two-dose-difference + SIT workflow, the modified workflow identified all putative NIF metabolites and improved the true-positive rate from 36.9 % to 71.0 % in the coincubation setup. Separate incubation yielded the most comprehensive profile (56 features) compared to coincubation (44) or mixed supernatants (38), indicating coincubation may obscure certain metabolites. Performance of the modified workflow remained consistent irrespective of dosing levels.

SIGNIFICANCE: By incorporating a mass shift defect filter into the two-dose-difference + SIT workflow, we more than doubled the true-positive discovery rate-from 36.9 % to 71.0 %-without sacrificing metabolite coverage. This streamlined, dose-independent method reduces false positives and accelerates reliable metabolite identification, offering a practical, resource-efficient platform for early-stage drug metabolism studies and mechanistic pharmacology.