Biological data on the cytotoxic and the ecotoxic effects of coated and uncoated nanoparticles were retrieved from the scientific literature. The mathematical treatment of these data was based on the use of perturbation theory (PT) operators. This enabled the development of a model that combined perturbation theory concepts with artificial neural networks (PTML-ANN). New nanoparticles not reported during the generation of the PTML-ANN model were used in a virtual screening experiment. For these new nanoparticles, the predictions performed by the PTML-ANN model converged with the experimental results.