This proposal contains a plan for a computer resource to support research using numerical simulation of biomedical systems. The principal objective is service: to provide national access to a computer system optimized for the computations of simulation--particularly those simulations requiring large amounts of computation. The philosophy of the proposal is to utilize commercially-availabel hardware and software as much as possible to reduce the startup time and development costs. Significant computing power will be obtained at modest cost by utilizing two special-purpose processors--one designed for rapid solution of systems of differential equations and the other for systems of algebraic equations (by vector and matrix calculations)--with a medium-sized minicomputer used as a host. Initial software support purchased for the system will include a dual operating system, two languages, and subroutine packages for numerical calculations and graphics as well as manufacturer-supplied software for the special-purpose processors. The proposal also includes research in the form of software development to support biomedical simulation, primarily modules or common functions to reduce programming effort, programs to assist in the selecting and viewing of output, and procedures for storing and accessing simulation results. Educational aspects of the proposal include written materials such as mannuals and a newsletter, short courses for novice and expert simulators, and the organization of symposia for the purpose of constructing broad, generally-applicable models. As an initial group of users of the resource, the proposal includes collaborative projects from a variety of basic biomedical sciences, including muscle physiology, neurophysiology, pharmacology, radiation physics, and toxicology. The objective of the resource is to make simulation more readily available to the biomedical research community, both by simplifying the necessary computer procedures and by providing access to a facility that usually cannot be afforded by single projects. We feel that the increased availability of simulation will have a substantial impact on biomedical research because simulation provides a necessary tool for thinking a bout complex systems, e.g. for effective experiment planning and analysis and for bringing together contributions from several disciplines to the understanding of a single process or system.
Showing the most recent 10 out of 42 publications
