Neural network modeling search for multitarget RAGE inhibitors with different target activity spectra

Five scaffolds (phenoxybenzene, benzofuran, azolobenzimidazole, oxoindole, and uracil), which are promising as sources of compounds with high RAGE inhibitory activity, were found in silico among ten structurally different classes of the synthesized substances using a multitarget neural network model of the dependence of the RAGE inhibitory activity on the affinity of the compounds to the target proteins of the RAGE—NF-κB signal pathway. The derivatives of these scaffolds are characterized by the predominantly post-receptor effect and manifest different target activity spectra to signal kinases of the RAGE—NF-κB chain. The revealed multitarget compounds were recommended for experimental investigation in order to determine their RAGE inhibitory activity. They can form a basis for the development of new promising drugs for the treatment of complications after diabetes mellitus and Alzheimer’s disease.