The goal of this project is to develop advanced mathematical, computational and visualization algorithms, software, methodology and tools which support the efficient application of computationally intensive science to key problems arising in the industrial sector. This work is a contribution of the National Institute of Standards and Technology to the Federal High Performance Computing and Communications program.
The projects undertaken by this task include the following.
Biological molecules (biomolecules) and processes are of great interest for adaptation to technologies such as chemical manufacturing, environmental remediation, and materials development. Modern genetic engineering provides methods to modify biomolecules, such as proteins, so that they have the desired properties for applications outside of living systems. This project is focussed on developing computational methods which relate the atomic structures of proteins to their properties and function.
In biological systems, chemical reactions are usually catalyzed by protein molecules called enzymes. Methods have been developed to allow the detailed study of enzymatic reactions using accurate computational models based on quantum mechanics. Quantum mechanical models are needed to describe the making and breaking of chemical bonds in a reaction that is under the control of an enzyme catalyst. Typical enzymes are very large molecules, containing thousands of atoms, for which quantum methods are intractable even on the largest computers. The new methods use a hybrid approach where the active portion of the enzyme, which is involved directly in the chemical reaction, is modeled with quantum mechanics while the bulk of the molecule and the solvent are treated by less expensive and more classical methods.
This hybrid method is being developed jointly with the research group of Professor Mark Gordon at Iowa State University (mark@si.fi.amelab.gov). The core of the computational approach is the GAMESS (Generalized Atomic and Molecular Structure System) quantum chemistry program which is maintained at Iowa State and distributed freely to hundreds of academic, government, and industrial research laboratories throughout the world. The GAMESS program has been adapted for parallel computing, which allows complex enzyme active sites to be modeled for the first time. Special modifications of GAMESS allow the interfacing of quantum and classical models in the study of complex molecular systems.
A more detailed project description is available.
Contact: Dr. Walter J. Stevens (walt@ibm1.carb.nist.gov)
The elucidation of protein structure plays a quintessential role in the development of our understanding of evolution and the function of biological processes.
Currently protein structures are determined experimentally by x-ray crystallography and NMR spectroscopy. These methods, although accurate, are time consuming and suffer from inherent drawbacks. There is a need for the development of theoretical methods which allow for the ab-initio calculation of protein structure.
It is important that the elucidation of protein structure can be performed at the same rate as sequencing of genomes. The phenotype (product) of a gene is a protein, and the pressure of natural selection is on the phenotype and not on the DNA. In order to be able to understand the meaning of large amounts of genomic data we need to understand how proteins (enzymes etc.) create specific functional structures.
The Protein Folding Problem refers to the combinatorial problems involved in enumerating the conformations of a given Protein molecule. Cyrus Levinthal (1968) outlined this in a simple paradox: Let each amino-acid residue in a 100 residue protein have 6 possible conformations, this leads to 6^100 possible conformations available for this protein, this calculation does not include sidechain conformations which will increase the number of degrees of freedom further. The question is now how does the protein fold given this large number of possible conformations. These simple calculations urge the development of new efficient and accurate search methods.
This project attempts to address the two main problems involved in the protein folding problem. The first problem is the understanding of the energetics involved in protein folding. This is here addressed from a thermodynamic view point, developing empirical Force Fields parameterised using experimentally determined protein structures. The second problem deals with the search problem outlined in Levinthals Paradox above.
The developed search methods and force-fields are applied to simulations of fragments of proteins which are known to contain structure, independently of the rest of the structure (Early Folding Units (Wetlaufer 1973)), and to small peptide chains that have been shown by nuclear molecular resonance (NMR) have a structure in solution.
A more detailed project description is available.
Contact: Dr. John Moult (jmoult@iris4.carb.nist.gov) or Dr. Jan T. Pedersen (jan@indigo7.carb.nist.gov)
The purpose of this task is to implement novel computational approaches to understanding the electronic properties of materials and the interaction of light and charged particles with matter.
Its primary customers are NIST laboratory programs that develop measurement methods needed for advancing electronic and optical technologies, in areas such as surface magnetism, soft x-ray optics, laser manipulation of atomic beams, and absolute radiometry.
We also work directly with commercial software developers, with the general goal of enhancing the reliability and efficiency of materials modeling applications.
Our principal technical focus is the use of computational parallelism on networked workstations, the environment that is of greatest relevance to most of our customers. The development of applications for such "virtual machines" has involved us in generic tool development in collaboration with colleagues in the NIST Computing and Applied Mathematics Laboratory.
Here is some information concerning ongoing projects:
Nanostructure electronics
Strong-field laser-matter interaction
NIST Parallel
Applications Development Environment
Optical manipulation of atomic beams
Electronic structure of large systems
NIST Physics Laboratory Home Page
The goal of this task is to develop mathematical algorithms, software, and related tools which increase the usability of high performance computing systems, and to investigate novel ways to disseminate this work to potential users. Three main subtasks are now underway.
The NIST Guide to Available Mathematical Software, a cross index and virtual repository of some 9,800 reusable components from more than 90 software packages, was recently made available for public access over the Internet. The system supports a variety mechanisms to help computational scientists and engineers locate and download appropriate software, centered about a detailed taxonomy of mathematical problems which has become a de facto standard. Users interact with information servers on the host gams.nist.gov using a variety of client programs, including World Wide Web (WWW) browsers (e.g., Mosaic) Gopher, and native GAMS clients. More than 10,000 users now connect to the GAMS system each month, representing over 100,000 individual client/server transactions. Contact: Dr. Ronald F. Boisvert (boisvert@nist.gov, 301-975-3812)
A suite of tools to manage distributed applications has been developed. These tools, the K-utilities, incorporate a variety of knowledge engineering techniques in in order to facilitate their use. Used in conjunction with a parallel computing system such as PVM, they can be used as a seamless, intelligent controller for coordinating applications or industrial processes across multiple platforms from PC's to supercomputers. The utilities have two interfaces: a command line interface, and the Parallel Applications Development Environment (PADE). PADE is being designed with the ability to learn and evolve over time to provide a better response to the user. Contact: Judith E. Devaney (judy@cam.nist.gov, 301-975-2882)
The encapsulation of efficient and robust mathematical algorithms in flexible and portable mathematical software packages is an important step in making high performance computers accessible for routine use in science and engineering. Recently, many application developers have begun to use the C++ and Fortran-90 programming languages to bring modern object-oriented software design methodology to their projects. Unfortunately, little mathematical software is available for such environments. Prototypes of several foundation software libraries and packages to support such applications are available.
NIST Computing and Applied Mathematics Laboratory Home Page
This HTML page is maintained by : Dr. Ronald F. Boisvert, boisvert@nist.gov, 301-975-3812.
This page was last updated on October 6, 1995.