Forefronts in large scale computing award: numerical simulation, parallel clusters, and the design of novel pharmaceutical agents for cancer treatment

Summary form only given. It is suggested that the application of numerical simulations and high-performance computers in rational drug design during the 1990s will become increasingly utilized in a manner analogous to the use of numerical simulations in other fields such as the aerospace and automotive engineering industries. One of the major bottlenecks on the discovery side of pharmaceutical development is the identification of high-quality, novel lead compounds which can be advanced into clinical trials. Because of the complexity and the interdependence of the data, the future will involve highly parallel environments to determine properties and further refinement of lead compounds in an expedient manner. This capability has been developed using a cluster of 20 IBM R/S 6000 systems which is capable of solving the range of critical algorithms pertinent to pharmaceutical discovery in parallel. The range of algorithms includes ab initio quantum mechanics, molecular dynamics, free energy perturbation, molecular shape analysis, graph theory, and molecular graphics.<
>