Integration of Per- and Polyfluoroalkyl Substance (PFAS) Fingerprints in Fish with Machine Learning for PFAS Source Tracking in Surface Water

Abstract

Per- and polyfluoroalkyl substances (PFASs) are a class of environmental contaminants that originate from various sources. The unique chemical fingerprints associated with many commercial products and industrial applications make PFASs ideal candidates for machine learning (ML)-assisted environmental forensics. Here, we propose a novel use of PFAS fingerprints in fish tissue from surface water systems to classify exposure from multiple sources of PFASs using a proof-of-concept demonstration. Three supervised ML classification techniques (k-nearest neighbors (KNN), decision trees, support vector machines) implementing two predictive features are used to classify literature-reported PFAS fingerprints in fish (n = 1057). The importance of additional predictive features was explored using brute force optimization of a multifeature KNN algorithm. The multiclass classification considered exposure to aqueous film-forming foam-impacted water, paper industry wastewater, diffuse sources, or PFASs undergoing long-range transport. The optimized classifiers demonstrated 85%–94% classification accuracy for this first known multiclass classification of PFASs for environmental forensics. The optimized classifiers also demonstrated 79%–92% classification accuracy with a set of independent external validation data (n = 192). Our results demonstrate that PFAS fingerprints in fish tissue may be an effective means of PFAS source tracking in surface water systems. The source code is provided for guidance on best practices for ML-assisted environmental forensics.

Multi-class machine learning classification of PFAS in environmental water samples: a blinded test of performance on unknowns

“…(ref. 4 and 11–16)). Some proposed methods have potential pitfalls, such as susceptibility to changing PFAS composition with transport or transformation of precursors, or potential challenges associated with detection limits, where specific compounds are too low in concentration to be detected in some samples.…”

“…As was done in previous work by the authors, 1–3 all component concentrations below detection limits were replaced with zeroes in both the training dataset and the dataset containing the unknowns, an approach that is essentially equivalent to placing all non-detects into a single bin for each component. For a full discussion of the justification for and implications of this approach, see Kibbey et al 1 Note that Stults et al 4 used substitution with a value related to the detection limit with success; it is likely that supervised machine learning classification is relatively insensitive to the handling of non-detects due to the fact that PFAS component concentrations often vary over orders of magnitude.…”

“…This hypothesis was strongly supported by the results of the work, which found that supervised machine learning exhibited great promise for distinguishing between AFFF and non-AFFF sources, even for difficult subsets of sample types. 2 Recent work by Stults et al 4 testing supervised machine learning for PFAS source identification in fish found similarly promising results for multiclass classification ( i.e. , distinguishing between PFAS from multiple source types).…”

Application Of Machine Learning To Understand Pfas Occurrence, Distribution, Transport And Removal In Water

“…Kwon et al, (2023) predicted bioactivities of PFAS. Other data sources include the United states environmental protection agency's water quality portal (USEPA) (Azhagiya Singam et al, 2020;DeLuca et al, 2023;Dong et al, 2023), PubChem Bioassay Database (Kwon et al, 2023), Pennsylvania Water quality network (Breitmeyer et al, 2023), from previously published studies on PFAS (Karbassiyazdi et al, 2022;Kibbey et al, 2020;Patel et al, 2022), lake and river data (Antell et al, 2023;Stults et al, 2023), Minnesota department of health (MDH) Government agency data (Breitmeyer et al, 2023;Fernandez et al, 2023;Li and Gibson, 2023) and experimental data (Cao et al, 2022;Sörengård et al, 2022;Wang et al, 2022). Some authors combined several public data for their machine learning predictions.…”

New PFASs Identified in AFFF Impacted Groundwater by Passive Sampling and Nontarget Analysis

“…This knowledge gap has led to the emergence of suspect screening and nontarget analysis (NTA) approaches, which aim to more comprehensively analyze the PFASs’ chemical space, including unanticipated and unknown compounds. − To date, the NTA methods have been used to characterize AFFF formulations, − PFASs in fire fighters’ blood, PFASs in cow blood, soil and concrete, landfill leachates, surface water, − and groundwater. , However, there is limited information about biotransformation, degradation products, environmental mobility, and persistence of PFASs in groundwater.…”