A systematically optimized two-dose differential strategy integrating stable isotope tracing and mass shift defect filtering for reducing false positives in high-resolution mass spectrometry-based drug metabolite profiling.

Drug metabolite detection using metabolomics-based approaches is often challenged by high false-positive rates and the limited availability of authentic reference standards. In this study, we systematically optimized a data processing workflow that integrates a two-dose differential strategy with stable isotope tracing (SIT) and mass shift defect filter (MSDF) to improve the detection and confirmation of drug-related metabolites. Using isotopically labeled (D0/D3) compounds, metabolite features were confirmed based on MS/MS fragmentation profiles and characteristic isotopic mass shifts, providing indirect yet robust evidence for metabolite assignment. A total of 56 sildenafil-related metabolite features were putatively detected following MS/MS-based confirmation. Comparative analysis of three incubation setups revealed that the separated incubation setup consistently yielded the largest number of metabolite features, despite showing a relatively modest improvement in detection rate after MSDF incorporation. Notably, mixing D0- and D3-labeled compounds within the same incubation tube resulted in a marked reduction in metabolite detection, consistent with previous findings and underscoring the importance of experimental design in isotope-assisted metabolomics studies. The effects of key analytical parameters, including MSDF threshold, sample size, and retention time tolerance, were systematically evaluated. An MSDF window with an absolute deviation of <0.12 Da and a retention time tolerance of 0.2 min, with a sample size of three paired samples, were identified as optimal settings that balance detection rate and metabolite coverage. Overall, this work demonstrates a robust and scalable workflow for comprehensive drug metabolite profiling and provides practical guidance for optimizing metabolomics-based metabolite identification strategies in the absence of authentic reference standards.