ToxFix aims to apply state-of-the-art computational models to advance predictive toxicology and design of safer chemicals in much the same way that computer simulations have advanced drug discovery in the past decades. Instead of deriving predictive parameters from chemical structure, which is commonplace in computational toxicology, our approach is to simulate chemical interactions and compute their energy landscapes, which we believe are more reflective of biological activity.
The cornerstone of ToxFix is a proprietary prediction software called CADRE (Computer Aided Discovery and REdesign), which combines quantum mechanical modeling, expert rules and multivariate statistics to predict toxicity of chemicals for several key endpoints. Prototypes of CADRE were developed under Sustainability A to Z, LLC with support from the National Science Foundation Small Business Innovation Research program and the Connecticut Innovations venture fund. Since then we have worked closely with academic and industrial partners to further develop and validate the CADRE methodology. For regulatory purposes, all models within CADRE comply with OECD principles for QSAR validation. Utilizing high-performance computing clusters, CADRE offers fast predictions with accuracyrivaling in vivo tests.
The ToxFix platform includes ToxSpec, a unique tool that predicts physicochemical properties, biological activity and biodegradation entirely from NMR spectra of pure substances or mixtures. The method achieves accuracy equal to or higher than the best structure-based estimation methods.