Dr. Luciana Teresa Dias Cappelini is a postdoctoral scholar at Florida International University working with Dr. Natalia Soares Quinete. This meeting will occur on Wednesday, September 23rd, from 12 pm – 1 pm EST. Please reach out if you did not receive the link through the newsletter and would like to attend.
Talk Title: “R-Based Pipeline for the Assignment of PFAS-Specific Schymanski Confidence Levels Transformation Products Generated by e-beam.”
Abstract: Per- and polyfluoroalkyl substances (PFAS) represent a class of persistent contaminants widely distributed in the environment due to their widespread use in commercial and industrial products and high stability of carbon–fluorine bonds. Among the technologies investigated for PFAS degradation, electron beam irradiation (E-beam) has emerged as a promising approach because of its ability to promote defluorination and structural transformation. However, the interpretation of data obtained by liquid chromatography high-resolution mass spectrometry (LC-HRMS) after e-beam treatment remains a significant challenge, mainly because of the large number of tentatively identified features detected and the difficulty of assigning reliable confidence levels. In non-targeted screening workflows, the assignment of confidence levels based on the Schymanski scale is generally performed manually, making the process labor-intensive, time-consuming, and susceptible to analyst-dependent interpretation. In this study, a pipeline was developed to automate the assignment of confidence levels to transformation products of 5:3FTCA, 6:3FTS, and 7:3FTCA generated after e-beam irradiation at doses of 250 and 500 kGy, at pH 12, and with low oxygen concentration. LC-HRMS data acquired using an Orbitrap IQ-X instrument (Thermo Scientific) were processed in Compound Discoverer v3.5, and the resulting feature tables were exported to a workflow implemented in R. The pipeline integrated exact mass, mass error, isotopic pattern, retention time, adduct information, and MS/MS fragmentation evidence to automatically classify each feature according to Schymanski’s confidence levels 1 to 5. Level 1 corresponds to confirmation with an authentic analytical standard, whereas levels 2 to 5 represent decreasing degrees of structural certainty based on spectral and analytical evidence. This workflow is currently being evaluated using samples obtained during the irradiation experiments to assess its applicability to the systematic interpretation of LC-HRMS features. By standardizing the assignment procedure, the proposed strategy reduces reliance on manual evaluation and improves the consistency and reproducibility of confidence level assignments. The results indicate that the pipeline is a promising tool for the automated assignment of PFAS confidence levels in non-targeted screening studies and may subsequently be evaluated using larger sample sets and additional PFAS classes.