New Biological Discoveries through Data Intensive Computing
November 08, 2004
RICHLAND, Wash. –
The field of biology is undergoing a revolution, transformed by sophisticated technologies from a qualitative, descriptive science to one that's quantitative and predictive. These technologies are producing a wealth of biological data that, once collected, analyzed and interpreted holistically, will form the basis for applications ranging from the development of cancer treatments to the creation of novel bioremediation technologies that will help clean up the worst Superfund sites.
In an effort to speed up the development of health and environmental remediation solutions, Pacific Northwest and Oak Ridge National Laboratories are teaming to identify areas of biological research that are in urgent need of computing capabilities. Known as "data-intensive computing," the work will extract knowledge from large data sources. Although a mouthful, data-intensive computing is really quite simple. It's using sophisticated computers to sort through mounds of data and present biologists with solutions in the form of graphics, scenarios, formulas, new hypotheses and more. Consistent with the laboratories' expertise, the project leads will focus on three specific areas of biology: proteomics, biomolecular simulations and biological network analysis.
PNNL and ORNL each received $1 million for the project, informally referred to as "BioPilot," from the Department of Energy's Office of Advanced Scientific Computing Research. BioPilot will be the first step for the two Department of Energy laboratories to develop software and computer infrastructures that will address the areas of biology determined to have the greatest need for data-intensive computing.
"We have truly moved into a time where high performance, data intensive computing is essential to most areas of basic biological research. Our success at analyzing the massive volumes of data will hopefully form the basis for new biological discoveries." said T. P. Straatsma, PNNL's Associate Division Director for Computational Biology and Bioinformatics.
"The ORNL team has identified large-scale biological computing problems in which new computer structures and memory will make a huge difference," said Nagiza Samatova, a research scientist at ORNL's Computational Biology Institute. "The team has been working hard to demonstrate that the predictive simulations of complex biological systems could be enabled by tight coupling of advanced mathematics, algorithms and hardware architectures. It is a big challenge and the team is ready for it."
PNNL and ORNL are world leaders in computing and computational sciences. PNNL's supercomputer in the Environment Molecular Sciences Laboratory is currently the ninth fastest in the world and has been instrumental in providing data that advances research in areas such as global climate change, biogeochemistry and DNA analysis.
At ORNL, DOE established the Center for Computational Sciences, a high-performance computing research center, in 1992. It's a designated user facility with several missions, including helping to solve grand challenges in science and engineering. Earlier this year, the Center was chosen by DOE to lead a partnership with a goal of building the world's most powerful supercomputer by 2007. The National Leadership Computing Facility to be housed at ORNL will pool the partnership's computational resources for a sustained capacity of 50 trillion operations per second (50 teraflops) and a peak capacity of more than 250 teraflops.
PNNL and ORNL have been developing and refining high performance computational analysis tools for government and commercial clients for more than 15 years. The knowledge and experience garnered from these projects will be the foundation for the laboratories to continue to develop new data intensive analysis principles in support of basic scientific research in biology and other domain sciences.
Tags: Energy, Environment, Fundamental Science, Computational Science, Operations, Climate Science, Environmental Remediation, Chemistry, Biology, Proteomics, Supercomputer