Because of the emergence of a whole range of new diseases that affect humans, as well as the increasing problems of resistance to certain existing drug families. The need to design effective drugs is very important. The development of drugs today is growing in the sense of not only finding new drugs but also reducing the time and the cost of this procedure. The work done during this study is the in silico design of heterocyclic molecules potentially interesting drug precursors in the treatment of certain bacterial infections or in chemotherapy. The employed strategy involves the use of computational methods based on different methods of bioinformatics and biostatistics (QSAR-2D, QSAR-3D, Molecular Docking and the Druglike or ADME-Toxic...) to better rationalize the design of new drug with specific structures, better activities and good affinities against better targets.