This proposal describes the plan for a computer resource to develop, support, and encourage numerical simulation as a tool for biomedical research. The general goals of the resource are to provide researchers with access to commercial simulation hardware and software where appropriate, to develop new software and techniques to make simulation more productive and more widely available, and to facilitate communication between researchers using computer simulation. The resource will continue to provide service to local and remote researchers: access (through a commercial data network) to special-purpose simulation and array processors, purchased in the initial grant period, for large simulations requiring extensive numerical calculations. The resource will also serve as a gateway to remote supercomputers for biomedical simulation. A second minicomputer host will be added to relieve the overloading of the present VAX-11/750 host, separating to individual machines the functions of software development and hosting the special- purpose processors. On the basis of experience in developing resource activities during the first three years of funding, more emphasis will be placed on software development research in the renewal period. In particular, a simulation environment will be developed based on graphics workstations, integrating new programs for building simulation models, for providing transparent access to different computers for solving model equations, and for manipulating simulation models. Finite state machines and cellular automata will be studied to develop new techniques for simulating systems with distributed state variables. Communication between researchers using simulation will be encouraged through common access to the resource's computer system and by organizing topical symposia. The resource will continue to provide education by teaching short courses for novice and experienced simulators, by developing tutorials in simulation, and by publishing a newsletter and other descriptive materials. The proposal includes collaborative projects from several basic biomedical research areas and ongoing projects that are contributing to the development of new software and techniques. The large increase in the quantity and quality of data from biomedical systems has created a severe problem for researchers attempting to understand these systems. We believe that making simulation more available and efficient will have a substantial effect on biomedical research because simulation is the most powerful tool available for thinking about complex systems, for organizing hypotheses, and for planning experiments.
Showing the most recent 10 out of 42 publications
