A New CSRML Structure-Based Fingerprint Method for Profiling and Categorizing Per- and Polyfluoroalkyl Substances
Ann M Richard1, Ryan Lougee2, Matthew Adams2
1Center for Computational Toxicology & Exposure, Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, Durham, North Carolina 27711, United States.
Chemical Research in Toxicology
|March 2, 2023
Summary
This study introduces novel PFAS-specific chemotypes (TxP_PFAS) for analyzing per- and polyfluoroalkyl substances (PFAS). These chemotypes enable automated, structure-based categorization of PFAS, improving chemical inventory analysis.
Area of Science:
- Environmental Chemistry
- Cheminformatics
- Toxicology
Background:
- Per- and polyfluoroalkyl substances (PFAS) encompass over 14,000 diverse chemical structures.
- Analyzing and categorizing the vast PFAS structure space is crucial for understanding their environmental occurrence and potential concerns.
- Current methods for profiling PFAS may not be efficient for large-scale analysis.
Purpose of the Study:
- To develop a novel set of PFAS-specific chemotypes (TxP_PFAS) for enhanced cheminformatic analysis.
- To create a computationally efficient and reproducible method for categorizing PFAS based on chemical structure.
- To demonstrate the utility of TxP_PFAS in profiling and categorizing the PFASSTRUCT inventory.
Main Methods:
- Utilized publicly available ToxPrint chemotypes and the ChemoTyper application.
- Developed 129 TxP_PFAS chemotypes, including modified bond-type ToxPrints and fluorinated chain/ring patterns.
- Applied TxP_PFAS to profile the PFASSTRUCT inventory and compare with expert-based PFAS categories (e.g., OECD Global PFAS list).
Main Results:
- Created a new PFAS-specific fingerprint set (TxP_PFAS) with 129 chemotypes, reducing chemotype counts by an average of 54% compared to ToxPrints.
- Demonstrated that TxP_PFAS can be visualized, filtered, and used to construct chemically intuitive, structure-based PFAS categories.
- Showcased TxP_PFAS's ability to recapitulate expert-based PFAS categories using clear, computationally implementable structure rules.
Conclusions:
- TxP_PFAS chemotypes provide a powerful tool for the computational analysis and categorization of large PFAS inventories.
- This approach facilitates harmonized PFAS categorization, improves communication, and enables more efficient, chemically informed exploration of PFAS.
- The developed method supports computational modeling and reduces the need for expert consultation in PFAS classification.
