The abstract is read at the publisher; the summary is JClub's.
A machine-learning semi-quantification pipeline for LC-HRMS PFAS in five Korean industrial WWTPs predicts response factors with GMFE 2.21 (vs 2.75 RT-only) and enables SNTS expansion.
- Why it matters: Conventional target-only monitoring misses emerging, precursor-related, and transformation-derived PFAS, limiting how completely industrial wastewater effluents can be characterized.
- What they did: A Random Forest model trained on 35 reference PFAS used SMILES-derived descriptors plus retention-time-based selection and fallback rules to estimate instrument response factors, then integrated SNTS tentatively identified PFAS (confidence 2a–3).
- The result: SNTS added 12 PFAS beyond the target list and contributed 0.2–33% of the estimated PFAS sum in effluents (mean 6.4–141.1 ng/L), with influent-to-effluent increases suggesting precursor transformation may occur in WWTPs.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.