Predicting Toxic Equivalence Factors from super(13)C Nuclear Magnetic Resonance Spectra for Dioxins, Furans, and Polychlorinated Biphenyls Using Linear and Nonlinear Pattern Recognition Methods

Models are developed for predicting toxic equivalence factors for dioxins, furans, and PCBs, which are based on a relationship between a molecule's super(13)C nuclear magnetic resonance spectrum and its toxicity as measured by its toxic equivalence factor. The models belong to a class of predictive models known as a spectrometry data activity relationship. The models use any unexpectedly high toxic equivalence factor predictions to flag congeners currently assigned zero values, but that might make significant contributions to the dioxin-like toxicity of environmental mixtures. Model development is detailed, and illustrative results are provided to suggest that the super(13)C nuclear magnetic resonance spectral data contain information that is more reflective of each molecule's biochemical properties than calculated potentials and molecular alignment assumptions used in quantitative structure activity relationship models.