What question did this study set out to answer?

The aim is to enhance PFAS characterization using a virtual mass spectral library generated by a neural network.

Enhanced Environmental PFAS Characterization Using a Virtual High-Resolution Mass Spectral Library Generated by Transfer Learning-Based Neural Network

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates improved PFAS characterization in environmental samples, suggesting enhanced monitoring potential.

The aim is to enhance PFAS characterization using a virtual mass spectral library generated by a neural network.
Developed a mass spectral library using 10,553 PFAS structures from U.S. EPA and NORMAN databases.
Applied NPFAS-MS to groundwater and aqueous film-forming foam (AFFF) samples.
Deployed NPFAS-MS as a web-based tool for structure-to-spectrum prediction and library searching.
The NPFAS-MS model achieved a top-1 recall of 71.1% compared to 42.1% to 55.4% from other models.
38 potential PFAS were identified in AFFF products and 40 in groundwater samples.
Enabled characterization of a wider variety of PFAS, including emerging contaminants like ultrashort-chain derivatives.

Chen et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark