FlashMRM: An Automated Platform for MRM Method Generation and Iterative Optimization Based on High-Resolution Spectra

High-throughput targeted analysis in exposomics relies on multiple reaction monitoring (MRM) using liquid chromatography-triple quadrupole mass spectrometry (LC-TQMS), yet its method development remains limited by dependence on chemical standards, labor-intensive parameter generation, and complex matrix interferences (INTF). Here, we introduce FlashMRM, a web-based platform for the automated generation and optimization of MRM parameters by leveraging high-resolution mass spectrometry (HRMS) databases and experimentally preacquired biosample HRMS data. The pseudotranslation from the large-scale HRMS database to the TQMS database (TQDB) was achieved through mass unit normalization, retention time prediction, and cross-instrument collision energy conversion algorithms. FlashMRM integrated Pseudo TQDB containing over 25,000 candidate analytes and INTF TQDB, which encompasses ∼44,000 potential interference MS features in biological matrices, utilizing a dual-weighted scoring model to balance the sensitivity and specificity. Leveraging a library of pesticides with highly diverse chemical structures as a representative model, FlashMRM generated transitions for 255 pesticides with sensitivity comparable to experimentally developed MRM transitions while increasing the specificity score from 0.60 to 0.68. At a spiked concentration of 10 ng/L in urine, the number of Top 5 detectable transitions increased from 194 in experiments to 233 using FlashMRM with optimized specificity weighting, enhancing trace-level detection under biomatrix interference. In the analysis of human urine samples, FlashMRM achieved high-confidence TQMS detection for 55 of 99 targeted exposure biomarkers. FlashMRM enables sensitive and accurate large-scale targeted screening on TQMS without standards and extensive preanalysis injections, integrating optimized scoring algorithms and built-in databases within a user-friendly web platform.