Oncogenic mutations of the ras protooncogene are found in 20% of most types of human cancer and greater than 50% of colon and pancreatic carcinomas. Recently, the three-dimensional structures of ras in its """"""""active"""""""" and """"""""inactive"""""""" conformational states have been solved by X-ray diffraction. Here we propose structure based, computer-aided strategy for the design of ras inhibitors for use as therapeutic agents in the treatment of cancer. Phase I of this strategy will result in the design of a larger panel (>50 compounds) of candidate ligands, the most promising of which will be synthesized and evaluated (both biologically and by X-ray diffraction) in phase II. Phase II and III will involve further rounds of design, synthesis and testing.
