This study presents a review of biodegradability modeling efforts including a detailed assessment of two models developed using an artificial intelligence based methodology. Validation results for these models using an independent, quality reviewed database, demonstrate that the models perform well when compared to another commonly used biodegradability model, against the same data. The ability of models induced by an artificial intelligence methodology to accommodate complex interactions in detailed systems, and the demonstrated reliability of the approach evaluated by this study, indicate that the methodology may have application in broadening the scope of biodegradability models. Given adequate data for biodegradability of chemicals under environmental conditions, this may allow for the development of future models that include such things as surface interface impacts on biodegradability for example.
Baker, J. R., Gamberger, D., Mihelcic, J. R., & Sabljić, A. (2004). Evaluation of artificial intelligence based models for chemical biodegradability prediction. Molecules. Molecular Diversity Preservation International. https://doi.org/10.3390/91200989