This is a pretty standard approach, when you combine multiple "artifacts." Look for MSCS (multiple solvent crystal structures) on the experimental side or computational solvent mapping on the theoretical side (software like FTMap).
"Structure-based design of submicromolar, biologically active inhibitors of trypanosomatid glyceraldehyde-3-phosphate dehydrogenase", PNAS 96, 4273-4278. GAPDH contains a cofactor and the adenosyl moiety of the cofactor was used as a starting point for the design of a very potent (and selective) inhibitor. There are later publications as well... Also note that since this concerns a neglected disease, no drug will ever emerge from that type of research (no profits to be made).