In the beginning stages of drug design, scientists need to screen drug candidates from a great number of molecules. The use of computers as a screening tool has attracted a great deal of attention among scientists for its potential to significantly reduce the total cost and time. For accurate screening, the chemical characteristics and docking behaviors of ligands and receptors must be represented precisely in the computers. To obtain these precise representations for ligands and receptors, scientists need to execute necessary tools and arrange their parameters with a trail-and-error manner. We consider that this kind of trial-and-error process is a bottleneck to the drug design process as a whole. We proposed a drug design system that shares the users' trial-and-error processes as a "trial set."
