Leveraging QSAR and q-RASAR Modeling to Predict PFAS Toxicity and Identify Structural Drivers Using Rodent
Melanie Rios
Co-Presenters: Individual Presentation
College: The Dorothy and George Hennings College of Science, Mathematics and Technology
Major: Biology
Faculty Research Mentor: Supratik Kar
Abstract:
Perfluoroalkyl substances (PFAS) are of growing concern due to their persistence, bioaccumulation, and potential toxicity, raising significant environmental and human health risks. In this study, QSAR and q-RASAR models were developed using oral LD50 toxicity datasets from mice (58 PFASs) and rat (50 PFASs) to assess PFAS toxicity. The q-RASAR models outperformed the QSAR models, demonstrating superior predictive power and reliability. Subsequently, the q-RASAR model was employed to predict toxicity values for an external set of 2,522 PFAS compounds lacking experimental data. This approach not only filled critical toxicity data gaps but also identified key structural features responsible for PFAS toxicity. The results highlight the efficacy of q-RASAR modeling in large-scale toxicity predictions and provide valuable insights for future risk assessments and regulatory evaluations.