Various techniques for rational drug design are presented in the paper. The methods are based
on a substitution of antipharmacophore atoms of the molecules of training dataset by new atoms and/or
group of atoms increasing the atomic bioactivity increments obtained from an SAR study. Furthermore, a
design methodology based on the genetic algorithm DesPot for discrete optimization and generation of
new drug candidate structures is described. Additionally, wide spectra of SAR approaches (3D/4D QSAR
interior and exterior-based methods – BiS, CiS, ConGO, CoMIn, high-quality docking method - ReDock)
using MERA force field and/or AlteQ quantum chemical method for correct prognosis of bioactivity and
the bioactive probability have been described. The design methods are implemented at
www.chemosophia.com web-site for online computational services.