Review structure- and dynamics-based computational design of anticancer drugs

Cancer is a class of highly complex diseases involving multiple genes and multiple cross-talks between signaling networks. Cancer cells may be developed from inherited defects or acquired damages of DNA. However, many cancers are resistant to treatment, and metastasis of cancers makes the disease even more intractable. Secondary malignancies are frequently observed after cancer chemotherapy. The call for more effective cancer therapy is obligatory. Using drug-cocktails that combine multiple anti-cancer agents working in different mechanisms has been a standard treatment of cancers to overcome the drug resistance problem. More recently, design of multiple ligands (may be more easily understood as "multiple target ligands"), i.e., single agents that target multiple biomolecules in a rational manner, receives increasing attention. For those who work on computational drug design, such tasks serve as new opportunities for achieving drugs with more effective pharmacological actions, in addition to designing compounds with better binding affinity, better selectivity, or to discovering compounds that can exert their actions allosterically. Some recent methodological developments on computational drug design are reviewed, and a few recent drug design efforts on a selected set of targets (topoisomerases, Ras proteins, protein kinases, and histone deacetylases) toward cancer treatment and cancer prevention are summarized.