Drug research: myths, hype and reality

Lack of success with early combinatorial chemistry and high-throughput screening approaches resulted from inappropriate compound selection. We are now aware that screening compounds should be either 'lead-like' or 'drug-like' and have the potential to be orally available. However, there is a growing tendency to misuse such terms and to overestimate their importance, and to overemphasize ADME problems in clinical failure. Sometimes, this goes hand-in-hand with an uncritical application of high-throughput in silico methods. Structure-based and computer-aided approaches can only be as good as the medicinal chemistry they are based on. The search for new drugs, especially in lead optimization, is an evolutionary process that is only likely to be successful if new methods merge with classical medicinal chemistry knowledge.