A mixed directed-undirected data structure for a parallel implementation of a domain decomposition algorithm

The choice of data structures influences the parallelization, efficiency and the manageability of a mesh refinement program. We introduce a mixed directed-undirected graph that combines both communication and scheduling needs. An inverted index is maintained for the directed graph to improve code performance and readability.This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under contract No. W-7405-Eng-48.     

Dispersion modeling by kinematic simulation: Cloud dispersion model

A new technique has been developed to compute mean and fluctuating concentrations in complex turbulent flows (tidal current near a coast and deep ocean). An initial distribution of material is discretized into any small clouds which are advected by a combination of the mean flow and large scale turbulence. The turbulence can be simulated either by kinematic simulation (KS) or direct numerical simulation. The clouds also diffuse relative to their centroids; the statistics for this are obtained from a separate calculation of the growth of individual clouds in small scale turbulence, generated by KS. The ensemble of discrete clouds is periodically re-discretized, to limit the size of the small clouds and prevent overlapping. The model is illustrated with simulations of dispersion in uniform flow, and the results are compared with analytic, steady state solutions. The aim of this study is to understand how pollutants disperses in a turbulent flow through a numerical simulation of fluid particle motion in a random flow field generated by Fourier modes. Although this homogeneous turbulent is rather a “simple” flow, it represents a building block toward understanding pollutant dispersion in more complex flow. The results presented here are preliminary in nature, but we expect that similar qualitative results should be observed in a genuine turbulent flow.     

Symmetric Markov chains in ℤd: How fast can they move?

Summary Consider a reversible Markov chain X _n which takes values in a subset of ^d. If the steps of the chain are uniformly bounded and the invariant measure satisfies a mild regularity condition, Varopoulos, Carne and Kesten have obtained estimates on \lambda n^{1/2} )$$ " align="middle" border="0"> which exhibit a Gaussian tail in but blow up as n. Following Kesten's approach we derive bounds which are uniform in n in some special cases. Our main result, however, is an example which shows that in general the estimates of Varopoulos, Carne and Kesten are essentially best possible.Research partially supported by an S.E.R.C. (U.K.) visiting fellowship and an operating grant from N.S.E.R.C. of Canada     

Determination of 1-vinyl-2-pyrrolidone in human serum by high-performance liquid chromatography

Summary A simple method is described which allows the quantification of 1-vinyl-2-pyrrolidone (NVP) in human serum. NVP is extracted from serum with diethylether and determined with HPLC/UV-detection. 1-Cyclohexyl-2-pyrrolidone serves as an internal standard. The detection limit is 0.1 mg/l. The method has shown that NVP can enter the organisms of workers occupationally exposed to this substance.     

Automated separation and measurement of radioxenon for the Comprehensive Test Ban Treaty

A fully automatic radioxenon sampler/analyzer (ARSA) has been developed and demonstrated for the collection and quantitative measurement of the four xenon radionuclides,^131mXe(11.9 d),^133mXe(2.2 d),¹³³Xe(5.2 d), and¹³⁵Xe(9.1 hr), in the atmosphere. These radionuclides are important signatures in monitoring for compliance to a Comprehensive Test Ban Treaty (CTBT). Activity ratios of these radionuclides permit source attribution. Xenon, continuously and automatically separated from the atmosphere, is automatically analyzed by electron-photon coincidence spectrometry providing a lower limit of detection of about 100 μBq/m³. The demonstrated detection limit is about 100 times better than achievable with reported laboratory-based procedures for the short-time collection intervals of interest.     

Trap centers in Au-implanted MOS structures

Trap centers in the Si-SiO₂ interface region of MOS structures doped by ion implantation of gold have been investigated using constant capacitance deep level transient spectroscopy (CC-DLTS). Gold doses of 10¹²–3 × 10¹³ cm^–2 were implanted into the back surface of the wafers and were then redistributed during a diffusion anneal for 30 min at 1100° or 900° C. Three Au-related trap levels have been observed in the interface region, which were attributed to the Au-donor (E _v +0.35 eV), the Au-acceptor (E _v +0.53 eV), and the Au-Fe complex (E _v +0.45 eV). The trap concentration profiles show that the Si-SiO₂ interface affects the Au concentration in a depth range of 1 m from the interface and that gettering of Au occurs at the interface. The interface state density is independent of the Au concentration at the interface even for concentrations of 10¹⁵ cm^–3.     

Fast determination of 13C NMR chemical shifts using artificial neural networks

Nine different artificial neural networks were trained with the spherically encoded chemical environments of more than 500000 carbon atoms to predict their 13C NMR chemical shifts. Based on these results the PC-program "C_shift" was developed which allows the calculation of the 13C NMR spectra of any proposed molecular structure consisting of the covalently bonded elements C, H, N, O, P, S and the halogens. Results were obtained with a mean deviation as low as 1.8 ppm; this accuracy is equivalent to a determination on the basis of a large database but, in a time as short as known from increment calculations, was demonstrated exemplary using the natural agent epothilone A. The artificial neural networks allow simultaneously a precise and fast prediction of a large number of 13C NMR spectra, as needed for high throughput NMR and screening of a substance or spectra libraries.     

An in silico ensemble method for lead discovery: decision forest

Recent progress in combinatorial chemistry and parallel synthesis has radically changed the approach to drug discovery in the pharmaceutical industry. At present, thousands of compounds can be made in a short period, creating a need for fast and effective in silico methods to select the most promising lead candidates. Decision forest is a novel pattern recognition method, which combines the results of multiple distinct but comparable decision tree models to reach a consensus prediction. In this article, a decision forest model was developed using a structurally diverse training data set containing 232 compounds whose estrogen receptor binding activity was tested at the U.S. Food and Drug Administration (FDA)'s National Center for Toxicological Research (NCTR). The model was subsequently validated using a test data set of 463 compounds selected from the literature, and then applied to a large data set with 57,145 compounds as a screening example. The results show that the decision forest method is a fast, reliable and effective in silico approach, which could be useful in drug discovery.