NBCR to offer biomedical researchers access to HPC and data management facilities

San Diego 19 May 1999 The United States National Biomedical Computation Resource (NBCR) integrates, develops, and delivers transparent access to HPC or high-performance computing and data management, with the end goal of making sophisticated biomedical technologies as easy to use as the Web. NBCR was recently awarded a five-year $5.9 million grant which will allow to continue its research efforts. The focus of NBCR is to unite biomedical research efforts that might not otherwise come in contact, according to Peter Arzberger, who is executive director of the San Diego Supercomputer Center (SDSC) as well as NBCR principal investigator. Together these projects allow researchers to achieve results which, separately, might be impossible to obtain.

Advertisement

The United States National Biomedical Computation Resource (NBCR) integrates, develops, and delivers transparent access to HPC or high-performance computing and data management, with the end goal of making sophisticated biomedical technologies as easy to use as the Web. NBCR was recently awarded a five-year $5.9 million grant which will allow to continue its research efforts. The focus of NBCR is to unite biomedical research efforts that might not otherwise come in contact, according to Peter Arzberger, who is executive director of the San Diego Supercomputer Center (SDSC) as well as NBCR principal investigator. Together these projects allow researchers to achieve results which, separately, might be impossible to obtain.

Starting its second five-year period on May 1st 1999, NBCR joins biomedical researchers from SDSC, the University of California San Diego, and the Scripps Research Institute in projects spanning biological databases, visualization, genetic sequence analysis, cardiac modelling, and also electron microscopy. NBCR is funded by the National Institutes of Health through a National Center for Research Resources grant. In addition, NBCR organizes teaching workshops and classes in order to further disseminate information about the project's resources. From July 21st to 22nd 1999, Dr. Kim Baldridge will lead the next NBCR workshop on the subject of "Fundamentals of Quantum Mechanical Methods for Biomedical Systems".

As the Web evolved during the first five years, NBCR was at the forefront of efforts to make high-performance computing resources accessible through a Web interface. Today, NBCR provides Web interfaces to six servers allowing biologists to take advantage of SDSC's powerful high-performance computing infrastructure to compare biological structures, compare genetic sequences, search the Internet for information related to a specific molecule, or perform other computations. The Web site has recently been redesigned to establish easier access to the different servers and to enable biomedical researchers to request time on resources allocated by NPACI, the National Partnership for Advanced Computational Infrastructure or initiate collaboration with NBCR researchers.

During the next five years of the NBCR effort, six project investigators are leading collaborative efforts to develop, extend, and integrate technologies for high-performance computing, transparent access, data management, and visualization. Dr. Kim Baldridge, SDSC computational chemist and UCSD associate adjunct professor, is leading an effort in the area of computational biochemistry. In this initiative, the high-performance computing techniques and tools of first-principles quantum mechanics will be applied to biomedical questions while the tools will be made accessible to biomedical researchers. This type of capabilities stand to benefit drug design, enzyme analysis, and the interpretation as well as prediction of detailed molecular structures and properties.

Michael Gribskov, SDSC computational biologist, leads a team of researchers developing molecular pattern recognition tools which are accessible through the Web to biological scientists. Pattern recognition tools enable biologists to generate maps between a molecule's genetic sequence and its function. This is based on comparisons to known molecules. Transparent supercomputing makes these computationally demanding comparisons accessible without the steep learning curve associated with computing on unfamiliar systems. Phil Bourne, who is SDSC computational biologist, associate adjunct professor of pharmacology at UCSD, and an adjunct professor at the Burnham Institute, is contributing SDSC's expertise in biological data representation and query to develop a data-management environment for biological modelling.

This initiative will develop two databases, a Mapping Protein Space database and a Biological Image Repository, as well as provide database technology to the other NBCR projects. Dr. Arthur Olson, a professor of molecular biology at the Scripps Research Institute, and a team of collaborators are developing a visualization environment for multi-scale biomolecular modelling. Indeed, the visualization of complex molecular assemblies at different resolutions is becoming critical to understanding the data and models being produced by the biomedical community. Olson's project will develop a component-based environment that can be adapted to a great variety of biological visualization needs.

Mark Ellisman, professor of neuro-sciences and bioengineering and director of the National Center for Microscopy and Imaging Research at UCSD, will lead a research team in the design of remotely accessible advanced tools for electron tomographic reconstruction and also large-scale storage of electron microscope tomography data. This project will provide the ability to examine structures not suitable for X-ray crystallography and which are too small for light microscopes. The research team will integrate tools of high-performance computing, transparent access, and data management capabilities. Andrew McCulloch, professor of bioengineering at UCSD, heads a team in developing 3-D heart models with Web-based interfaces, allowing them to be visualized, fit to medical image data, or used to generate grids for computer modelling. Such models will depend on transparent access to high-performance parallel computers with integrated visualization and database management tools.


Leslie Versweyveld

[Medical IT News][Calendar][Virtual Medical Worlds Community][News on Advanced IT]