Revisiting proportion estimators

Proportion estimators are quite frequently used in many application areas. The conventional proportion estimator (number of events divided by sample size) encounters a number of problems when the data are sparse as will be demonstrated in various settings. The problem of estimating its variance when sample sizes become small is rarely addressed in a satisfying framework. Specifically, we have in mind applications like the weighted risk difference in multicenter trials or stratifying risk ratio estimators (to adjust for potential confounders) in epidemiological studies. It is suggested to estimate p using the parametric family (see PDF for character) and p(1 - p) using (see PDF for character), where (see PDF for character). We investigate the estimation problem of choosing c 0 from various perspectives including minimizing the average mean squared error of (see PDF for character), average bias and average mean squared error of (see PDF for character). The optimal value of c for minimizing the average mean squared error of (see PDF for character) is found to be independent of n and equals c = 1. The optimal value of c for minimizing the average mean squared error of (see PDF for character) is found to be dependent of n with limiting value c = 0.833. This might justifiy to use a near-optimal value of c = 1 in practice which also turns out to be beneficial when constructing confidence intervals of the form (see PDF for character).

Hepatozoon griseisciuri in grey squirrels (Sciurus carolinensis): changes of blood leucocyte numbers resulting from infection

Numbers of leucocytes in squirrels with gametocytes of Hepatozoon in their blood (infected) were compared with animals without gametocytes (uninfected). Typical values for leucocytes/mm(3) blood in uninfected squirrels were: leucocytes 5-7 x 10(3), granulocytes 3-4 x 10(3), lymphocytes 2-0 x 10(3) and monocytes 0-3 x 10(3) cells. Infection caused an increase in monocytes, lymphocytes and granulocytes, and there was a significant positive association between parasitaemia level and numbers of both total leucocytes and monocytes. Infected animals had more uninfected monocytes/mm(3) blood than did uninfected animals. The proportions of monocytes were more variable over time in infected animals, but no shift between infected and uninfected status was detected. Transfer of serum from infected squirrels to mice resulted in elevated counts of total blood leucocytes, monocytes and granulocytes, but not of lymphocytes, as compared with controls. Serum from squirrels with high parasitaemias had a more marked effect than serum from squirrels with low parasitaemias. Results indicate an infection - related monocytosis, possibly controlled by cytokines, that increases the number of cells available for invasion by gametocytes, thus enhancing the chances of parasite transmission.

A comparison of the reactivity of germylene and dimethylgermylene with some methylgermanes. Direct kinetic and quantum chemical studies

Time-resolved studies of germylene, GeH2, and dimethygermylene, GeMe2, generated by the 193 nm laser flash photolysis of appropriate precursor molecules have been carried out to try to obtain rate coefficients for their bimolecular reactions with dimethylgermane, Me2GeH2, in the gas-phase. GeH2 + Me2GeH2 was studied over the pressure range 1-100 Torr with SF6 as bath gas and at five temperatures in the range 296-553 K. Only slight pressure dependences were found (at 386, 447 and 553 K). RRKM modelling was carried out to fit these pressure dependences. The high pressure rate coefficients gave the Arrhenius parameters: log(A/cm(3) molecule(-1)s(-1)) = -10.99 +/- 0.07 and E-a = -(7.35 +/- 0.48) kJ mol(-1). No reaction could be found between GeMe2 + Me2GeH2 at any temperature up to 549 K, and upper limits of ca. 10(-14) cm(3) molecule(-1)s(-1) were set for the rate coefficients. A rate coefficient of (1.33 +/- 0.04) x 10(-11)cm(3) molecule(-1)s(-1) was also obtained for GeH2 + MeGeH3 at 296 K. No reaction was found between GeMe2 and MeGeH3. Rate coefficient comparisons showed, inter alia, that in the substrate germane Me-for-H substitution increased the magnitudes of rate coefficients significantly, while in the germylene Me-for-H substitution decreased the magnitudes of rate coefficients by at least four orders of magnitude. Quantum chemical calculations (G2(MP2,SVP)// B3LYP level) supported these findings and showed that the lack of reactivity of GeMe2 is caused by a positive energy barrier for rearrangement of the initially formed complexes. Full details of the structures of intermediate complexes and the discussion of their stabilities are given in the paper.

Surprisingly slow reaction of dimethylsilylene with dimethylgermane: time-resolved kinetic studies and related quantum chemical calculations

Time-resolved studies of silylene, SiH2, and dimethylsilylene, SiMe2, generated by the 193 nm laser flash photolysis of appropriate precursor molecules have been carried out to obtain rate constants for their bimolecular reactions with dimethylgermane, Me2GeH2, in the gas phase. SiMe2 + Me2GeH2 was studied at five temperatures in the range 299-555 K. Problems of substrate UV absorption at 193 nm at temperatures above 400 K meant that only three temperatures could be used reliably for rate constant measurement. These rate constants gave the Arrhenius parameters log(A/cm(3) molecule(-1) s(-1)) = -13.25 +/- 0.16 and E-a = -(5.01 +/- 1.01) kJ mol(-1). Only room temperature studies of SiH2 were carried out. These gave values of (4.05 +/- 0.06) x 10(-10) cm(3) molecule(-1) s(-1) (SiH2 + Me2GeH2 at 295 K) and also (4.41 +/- 0.07) x 10(-10) cm(3) molecule(-1) s(-1) (SiH2 + MeGeH3 at 296 K). Rate constant comparisons show the surprising result that SiMe2 reacts 12.5 times slower with Me2GeH2 than with Me2SiH2. Quantum chemical calculations (G2(MP2,SVP)//B3LYP level) of the model Si-H and Ge-H insertion processes of SiMe2 with SiH4/MeSiH3 and GeH4/MeGeH3 support these findings and show that the lower reactivity of SiMe2 with Ge-H bonds is caused by a higher secondary barrier for rearrangement of the initially formed complexes. Full details of the structures of intermediate complexes and the discussion of their stabilities are given in the paper. Other, related, comparisons of silylene reactivity are also presented.

IR spectrum and radiative forcing of CF4 revisited

Carbon tetrafluoride (CF4) is included as a greenhouse gas within the Kyoto Protocol. There are significant discrepancies in the reported integrated infrared (IR) absorption cross section of CF4 leading to uncertainty in its contribution to climate change. To reduce this uncertainty, the IR spectrum of CF4 was measured in two different laboratories, in 0 933 hPa of air diluent at 296 +/- 2K over the wavelength range 600-3700 cm(-1) using spectral resolutions of 0.03 or 0.50 cm(-1). There was no discernable effect of diluent gas pressure or spectral resolution on the integrated IR absorption, and a value of the integrated absorption cross section of (1.90 +/- 0.17) x 10(-16) cm(2) molecule(-1) cm(-1) was derived. The radiative efficiency (radiative forcing per ppbv) and GWP (relative to CO2) of CF4 were calculated to be 0.102 W m(-2) ppbv(-1) and 7200 (100 year time horizon). The GWP for CF4 calculated herein is approximately 30% greater than that given by the Intergovernmental Panel on Climate Change (IPCC) [ 2002] partly due to what we believe to be an erroneously low value for the IR absorption strength of CF4 assumed in the calculations adopted by the IPCC. The radiative efficiency of CF4 is predicted to decrease by up to 40% as the CF4 forcing starts to saturate and overlapping absorption by CH4, H2O, and N2O in the atmosphere increases over the period 1750-2100. The radiative forcing attributable to increased CF4 levels in the atmosphere from 1750 to 2000 is estimated to be 0.004 W m(-2) and is predicted to be up to 0.033 W m(-2) from 2000 to 2100, dependent on the scenario.

Macrocyclic amines as structure-directing agents for the synthesis of three-dimensional antimony-sulfide frameworks

A new family of antimony sulfides, incorporating the macrocyclic tetramine 1,4,8,11-tetraazacyclotetradecane ( cyclam), has been prepared by a hydrothermal method. [C10N4H26][Sb4S7] (1), [Ni(C10N4H24)][Sb4S7] (2), and [Co(C10N4H24)](x)[C10N4H26](1-x)[Sb4S7] (0.08 <= x <= 0.74) (3) have been characterized by single-crystal X-ray diffraction, elemental analysis, thermogravimetry, and analytical electron microscopy. All three materials possess the same novel three-dimensional Sb4S72- framework, constructed from layers of parallel arrays of Sb4S84- chains stacked at 90 to one another. In 1, doubly protonated macrocyclic cations reside in the channel structure of the antimonysulfide framework. In 2 and 3, the cyclam acts as a ligand, chelating the divalent transition- metal cation. Analytical and X-ray diffraction data indicate that the level of metal incorporation in 2 is effectively complete, whereas in 3, both metalated and nonmetalated forms of the macrocycle coexist within the structure.

Silica-encapsulated nanomagnetic particle as a new recoverable biocatalyst carrier

Enzymes are versatile biocatalysts with major advantages of ultrahigh reaction selectivity and specificity under mild conditions, which currently find increasing applications. However, their applications are often hampered by difficulties in recovery and recycling. As a result, we carried out detailed investigations on the synthesis and characterization of silica-encapsulated iron oxide magnetic nanoparticles of controlled dimension as an enzyme carrier. It is shown that the relatively smaller sized silica-coated magnetic nanoparticle prepared by the microemlusion technique can a carry bulky enzyme, beta-lactamase, via chemical linkages on the silica overlayer without severely blocking the enzymatic active center ( which is commonly encountered in conventional solid supports). An activity study by Michalis-Menten kinetics reflects that this new type of immobilization allows enzyme isolation with accessibility as good as free enzyme. The recovery and reusability of the nanoparticle-supported enzyme upon application of magnetic separation are also demonstrated.

Preparation of nanomagnetic absorbent for partition coefficient measurement

In this paper, we report a new method based on supercritical carbon dioxide (scCO(2)) to fill and distribute the porous magnetic nanoparticles with n-octanol in a homogeneous manner. The high solubility of n-octanol in scCO(2) and high diffusivity and permeability of the fluid allow efficient delivery of n-octanol into the porous magnetic nanoparticles. Thus, the n-octanol-loaded magnetic nanoparticles can be readily dispersed into aqueous buffer (pH 7.40) to form a homogenous suspension consisting of nano-sized n-octanol droplets. We refer this suspension as the n-octanol stock solution. The n-octanol stock solution is then mixed with bulk aqueous phase (pH 7.40) containing an organic compound prior to magnetic separation. The small-size of the particles and the efficient mixing enable a rapid establishment of the partition equilibrium of the organic compound between the solid supported n-octanol nano-droplets and the bulk aqueous phase. UV-vis spectrophotometry is then applied to determine the concentration of the organic compound in the aqueous phase both before and after partitioning (after magnetic separation). As a result, log D values of organic compounds of pharmaceutical interest determined by this modified method are found to be in excellent agreement with the literature data. (c) 2006 Elsevier B.V. All rights reserved.

Evaluation of milk urea nitrogen as a diagnostic of protein feeding

An evaluation of milk urea nitrogen (MUN) as a diagnostic of protein feeding in dairy cows was performed using mean treatment data (n = 306) from 50 production trials conducted in Finland (n = 48) and Sweden (n = 2). Data were used to assess the effects of diet composition and certain animal characteristics on MUN and to derive relationships between MUN and the efficiency of N utilization for milk production and urinary N excretion. Relationships were developed using regression analysis based on either models of fixed factors or using mixed models that account for between-experiment variations. Dietary crude protein (CP) content was the best single predictor of MUN and accounted for proportionately 0.778 of total variance [ MUN (mg/dL) = -14.2 + 0.17 x dietary CP content (g/kg dry matter)]. The proportion of variation explained by this relationship increased to 0.952 when a mixed model including the random effects of study was used, but both the intercept and slope remained unchanged. Use of rumen degradable CP concentration in excess of predicted requirements, or the ratio of dietary CP to metabolizable energy as single predictors, did not explain more of the variation in MUN (R-2 = 0.767 or 0.778, respectively) than dietary CP content. Inclusion of other dietary factors with dietary CP content in bivariate models resulted in only marginally better predictions of MUN (R-2 = 0.785 to 0.804). Closer relationships existed between MUN and dietary factors when nutrients (CP to metabolizable energy) were expressed as concentrations in the diet, rather than absolute intakes. Furthermore, both MUN and MUN secretion (g/d) provided more accurate predictions of urinary N excretion (R-2 = 0.787 and 0.835, respectively) than measurements of the efficiency of N utilization for milk production (R-2 = 0.769). It is concluded that dietary CP content is the most important nutritional factor influencing MUN, and that measurements of MUN can be utilized as a diagnostic of protein feeding in the dairy cow and used to predict urinary N excretion.

Nonlinear system identification using particle swarm optimisation tuned radial basis function models

A novel particle swarm optimisation (PSO) tuned radial basis function (RBF) network model is proposed for identification of non-linear systems. At each stage of orthogonal forward regression (OFR) model construction process, PSO is adopted to tune one RBF unit's centre vector and diagonal covariance matrix by minimising the leave-one-out (LOO) mean square error (MSE). This PSO aided OFR automatically determines how many tunable RBF nodes are sufficient for modelling. Compared with the-state-of-the-art local regularisation assisted orthogonal least squares algorithm based on the LOO MSE criterion for constructing fixed-node RBF network models, the PSO tuned RBF model construction produces more parsimonious RBF models with better generalisation performance and is often more efficient in model construction. The effectiveness of the proposed PSO aided OFR algorithm for constructing tunable node RBF models is demonstrated using three real data sets.

Identification of nonlinear systems using generalized kernel models

Nonlinear system identification is considered using a generalized kernel regression model. Unlike the standard kernel model, which employs a fixed common variance for all the kernel regressors, each kernel regressor in the generalized kernel model has an individually tuned diagonal covariance matrix that is determined by maximizing the correlation between the training data and the regressor using a repeated guided random search based on boosting optimization. An efficient construction algorithm based on orthogonal forward regression with leave-one-out (LOO) test statistic and local regularization (LR) is then used to select a parsimonious generalized kernel regression model from the resulting full regression matrix. The proposed modeling algorithm is fully automatic and the user is not required to specify any criterion to terminate the construction procedure. Experimental results involving two real data sets demonstrate the effectiveness of the proposed nonlinear system identification approach.

Global exponential periodicity of a class of bidirectional associative memory networks with finite distributed delays

In this paper, we study the periodic oscillatory behavior of a class of bidirectional associative memory (BAM) networks with finite distributed delays. A set of criteria are proposed for determining global exponential periodicity of the proposed BAM networks, which assume neither differentiability nor monotonicity of the activation function of each neuron. In addition, our criteria are easily checkable. (c) 2005 Elsevier Inc. All rights reserved.