Science Inventory

Establishing QA/QC Protocols and Tools to Support Non-targeted Analyses of Environmental Matrices

Citation:

Sayre-Smith, N., L. Brunelle, A. Batt, A. Chao, E. Carr, T. Cathey, J. McCord, J. Minucci, Tom Purucker, A. Rashid, D. Smith, A. Williams, D. Alvarez, E. Furlong, S. Glassmeyer, D. Kolpin, M. Mills, AND J. Sobus. Establishing QA/QC Protocols and Tools to Support Non-targeted Analyses of Environmental Matrices. ISES, Chicago, IL, August 27 - 31, 2023. https://doi.org/10.23645/epacomptox.23792595

Impact/Purpose:

N/A

Description:

Non-targeted analysis (NTA) utilizing high-resolution mass spectrometry is an increasingly popular tool for environmental science given its ability to facilitate exploratory exposomic investigations and identify contaminants of emerging concern. Generalized reporting standards, like those promulgated in the NTA “Study Reporting Tool” (SRT), are now available for NTA practitioners. Yet, wide-spread standardization of NTA procedures is ongoing, and practical demonstrations of best-practices (as per the NTA SRT) remain lacking. Addressing this need, we report a proof-of-concept analysis in which a combination of vendor software and a prototype U.S. Environmental Protection Agency (USEPA) web application were used to generate critical QA/QC outputs for an environmental NTA study. NTA data used for this demonstration were from a study by the U.S. Geological Survey and USEPA to investigate the fate and transport of water contaminants throughout a watershed representative of typical de facto water reuse. We explored NTA QA/QC as it relates to: (1) Study Design (involving pooled samples, method blanks, tracer spikes, and system-suitability mixtures [SSMs]), (2) Data Acquisition (involving technical replicates and randomized run sequences), and (3) Data Outputs (including measures of accuracy, precision, sensitivity, and reproducibility). Physical outputs of the QA/QC workflows include summary tables of tracer and SSM data, run sequence plots demonstrating batch and matrix effects, and several visualizations (e.g., logic trees and heat maps) summarizing results of data filtering and flagging steps. After ensuring data quality, “clean” features are further examined to probe spatial and temporal trends, and direct prioritization efforts for compound annotation and follow-up confirmation.