An in silico screening study and design of potent cognition agents

The binding affinities of human histamine H3 antagonists were predicted using an enhancement replacement method - partial least squares (ERM-PLS) model and the results were compared with those of genetic algorithm (GA)-PLS and stepwise linear regression (SW)-PLS models. Based on this accurate, robust and reliable ERM-PLS model, an in silico screening study was performed, which resulted in some new and potent cognition agents, some of which show equal or even better affinity and ADMET properties than the previous structures. The effects of structural moieties on the activity of the compounds were discussed.