Abstract

The best – known software applications for 3D visualization and evaluation of bioisosteric molecules will be compared. The software applications are based on parameters, which can be divided mainly into two directions – emphasizing electrostatic properties, wave functions and presence of electrons in different energy states, and QSAR – based – quantitative methods – empirical examination of molecules, which use mainly statistical data collection, thus evaluating the properties of molecules through regression models to investigate similarity. The first group of innovative methods are superior, because of their analytical purposes that are not mining blindly for bioisosterism through empirical statistical data, but analytically detect similarities according to molecules’ shapes. In conclusion – the use of GPUs for parallel processing of matrix-based operation – for the calculations of quantum vectors – would lead to repeated acceleration of the speed of these calculations. Finally, this study demonstrates the GPU as another factor, which is important for the improvement of the evaluation methods of studying and visualization of molecular similarity.