A lower bound on the performance of the quadratic discriminant function

The quadratic discriminant function is often used to separate two classes of points in a multidimensional space. When the two classes are normally distributed, this results in the optimum separation. In some cases however, the assumption of normality is a poor one and the classification error is increased. The current paper derives an upper bound for the classification error due to a quadratic decision surface. The bound is strict when the class means and covariances and the quadratic discriminant surface satisfy certain specified symmetry conditions.     

Synthesis of a new class of uridine phosphorylase inhibitors

A new series of potent uridine phosphorylase inhibitors have been prepared from barbituric acid. Among them, 1-[(2-hydroxyethoxy)methyl]-5-)(m--benzyloxy)benzylbarbituric acid ( 37 , BBBA) is the most promising having a K_i value of 1.1 ± 0.2 nM with uridine phosphorylase from human liver. The new inhibitors are easily synthesized and are better inhibitors of human uridine phosphorylase than their uracil counterparts.     

A semiclassical reactive flux method for the calculation of condensed phase activated rate constants

A semi classical reactive flux algorithm for calculating thermally activated rate constants is presented which is based on a semi-classical transition state theory due to Chapman, Garrett and Miller [J. Chem. Phys. 63 (1975) 2710]. This reactive flux technique, when combined with the semiclassical TST, enables one to describe dynamical recrossings of the transition state on the same footing as tunneling effects. Most importantly, the method is readily applied to nonlinear multidimensional systems over a wide range of temperatures. It will be shown that the method works very well for a variety of existing models.     

On estimating pressure build-up in severely accelerated,shocked flow

A new approach to the calculation of the high pressures characterizing the flow field in front of a piston undergoing severe acceleration over the short term is presented. In contrast with previous approaches where the computational domain is altered but which stop short of transforming velocities, here the problem is solved in an accelerating non-Euclidean co-ordinate system where the piston is stationary. The method is applied to a study of the problem of premature sabot separation. Through use of Harten's second-order-accurate TVD scheme, flow simulations are performed for both 1D and 3D axisymmetric geometries. The simple 1D model gives pressure profiles surprisingly close to those of the more physical 3D model.     

On the Role of Mixed Strategies in Elementary Decision Analysis and Related Decision-Support-System Treatments

We first demonstrate that mixed strategies are relevant in decision analysis for a maximin decision-maker quite apart from any game-theory considerations. This rectifies the apparent misconception that results from MS/OR textbooks which discuss mixed strategies only in the game-theory setting. Next we show an example of an implementable mixed strategy, by which we mean a mixed decision strategy which does not require randomization for its implementation. This application is to portfolio construction.     

The ICP-MS approach to environmental studies

The certification of marine materials for trace metals is a process which challenges every technique involved, especially if a technique is as recent as inductively coupled plasma mass spectrometry (ICP-MS), Developmental work was required for several materials (natural waters, biological materials, marine sediments). It is reviewed here, in an attempt to show how one can take full advantage of ICP-MS. This includes a review of the digestion procedures developed for the multielement analysis of biological materials and marine sediments in order to minimize spectroscopic interferences. The multielement analysis of natural waters is also reviewed, in particular that of saline waters which requires a separation of the analytes from the alkali and alkaline earths elements and a preconcentration of the analytes on a column of silicaimmobilized 8-hydroxyquinoline. The potential of performing this separation/preconcentration procedure on-line is showed using both published and original results. Finally, the application of ICP-MS to speciation is illustrated by the determination of methylmercury in biological materials after extraction, and by the determination of arsenic species by high performance liquid chromatography coupled to ICP-MS.Presented in part at the 1989 European Winter Conference on Plasma Spectrochemistry, Reutte, AustriaThe work described was carried out while the author was a Research Associate at the Analytical Chemistry Section, Chemistry Division, National Research Council of Canada, Ottawa, K1A OR9, Canada     

Kinetic and structural characterization of human manganese superoxide dismutase containing 3-fluorotyrosines

Incorporation of 3-fluorotyrosine and site-specific mutagenesis have been used with stopped-flow spectrophotometry and pulse radiolysis to investigate the catalytic properties of human manganese superoxide dismutase (MnSOD). All of the nine tyrosine residues in each of the four subunits of the homotetramer of human MnSOD were replaced with 3-fluorotyrosine. Previous studies showed that the crystal structures of the unfluorinated and fluorinated human MnSOD are nearly superimposable with the root-mean-square deviation for 198 -carbon atoms at 0.3 Å. However, the catalytic activity k_cat/K_m of the fluorinated MnSOD at 30 μM⁻¹ s⁻¹ was less than unfluorinated wild type at 800 μM⁻¹ s⁻¹. Comparison of the values of k_cat/K_m for fluorinated and unfluorinated wild-type andY34F MnSOD showed that this decrease for the fluorinated enzyme was in significant part due to 3-fluorotyrosine residues distant (>7 Å) from the active-site metal, not to 3-fluorotyrosine at position 34 close (5 Å) to the metal. Although many rate constants for the catalysis are decreased by this fluorination, the rate of dissociation of the product-inhibited complex appears unchanged by the presence of fluorinated tyrosines. These results suggest that Tyr34 is not a proton donor in the release of the product-inhibited complex, which involves protonation of a peroxo complex of the metal with release of hydrogen peroxide.