The behaviour of the multi-layer perceptron and the support vector regression learning methods in the prediction of NO and NO2 concentrations in Szeged, Hungary

The main aim of this paper is to predict NO and NO₂ concentrations 4 days in advance by comparing two artificial intelligence learning methods, namely, multi-layer perceptron and support vector machines, on two kinds of spatial embedding of the temporal time series. Hourly values of NO and NO₂ concentrations, as well as meteorological variables were recorded in a cross-road monitoring station with heavy traffic in Szeged, in order to build a model for predicting NO and NO₂ concentrations several hours in advance. The prediction of NO and NO₂ concentrations was performed partly on the basis of their past values, and partly on the basis of temperature, humidity and wind speed data. Since NO can be predicted more accurately, its values were considered primarily when forecasting NO₂. Time series prediction can be interpreted in a way that is suitable for artificial intelligence learning. Two effective learning methods, namely, multi-layer perceptron and support vector regression are used to provide efficient non-linear models for NO and NO₂ time series predictions. Multi-layer perceptron is widely used to predict these time series, but support vector regression has not yet been applied for predicting NO and NO₂ concentrations. Three commonly used linear algorithms were considered as references: 1-day persistence, average of several day persistence and linear regression. Based on the good results of the average of several day persistence, a prediction scheme was introduced, which forms weighted averages instead of simple ones. The optimization of these weights was performed with linear regression in linear case and with the learning methods mentioned in non-linear case. Concerning the NO predictions, the non-linear learning methods give significantly better predictions than the reference linear methods. In the case of NO₂, the improvement of the prediction is considerable, however, it is less notable than for NO.     

Determining the minimum grain size in severe plastic deformation process via first-principles calculations

Although the stacking fault energy (SFE) is a fundamental variable determining the minimum grain size (d_min) obtainable in severe plastic deformation (SPD) processes, its accurate measurement is difficult. Here we establish the SFEs of binary Pd–Ag, Pd–Cu, Pt–Cu and Ni–Cu solid solutions using the axial interaction model and the supercell model in combination with first-principles theory. The two models yield consistent formation energies. For Pd–Ag, Pd–Cu and Ni–Cu, the theoretical SFEs agree well with those from the experimental measurements. For Pt–Cu no experimental results are available, and thus our calculated SFEs represent the first reasonable predictions. We discuss the correlation of the SFE and d_min in SPD experiments and show that the d_min values can be evaluated from first-principles calculations.     

Analysis of Modified Nucleotide Aptamer Library Generated by Thermophilic DNA Polymerases

One of the pivotal steps in aptamer selection is the amplification of target-specific oligonucleotides by thermophilic DNA polymerases; it can be a challenging task if nucleic acids possessing modified nucleotides are to be amplified. Hence, the identification of compatible DNA polymerase and modified nucleotide pairs is necessary for effective selection of aptamers with unnatural nucleotides. We present an in-depth study of using 5-indolyl-AA-dUTP (TAdUTP) to generate oligonucleotide libraries for aptamer selection. We found that, among the eight studied DNA polymerases, only Vent(exo-) and KOD XL are capable of adapting TAdUTP, and that replacing dTTP did not have a significant effect on the productivity of KOD XL. We demonstrated that water-in-oil emulsion PCR is suitable for the generation of aptamer libraries of modified nucleotides. Finally, high-throughput sequence analysis showed that neither the error rate nor the PCR bias was significantly affected by using TAdUTP. In summary, we propose that KOD XL and TAdUTP could be effectively used for aptamer selection without distorting the sequence space of random oligonucleotide libraries.     

Influence of silicon doping on the nanomorphology and surface chemistry of a wood-based carbon molecular sieve

Carbon–silica molecular sieves were prepared by carbonization of Scotch fir (Pinus sylvestris) after impregnation with aqueous waterglass (Na_xSi_yO_z, where x, y and z may take a range of values). Compared to Si-free samples, doping significantly modifies the structure that forms during the carbonization process. For carbonization temperatures between 600 °C and 1000 °C, doped samples shrink less than undoped samples, indicating increased mechanical strength. The specific surface area and pore volume develop in a combined self-activation and chemical vapour deposition (CVD) process. Nevertheless, the presence of the sodium silicates limits self-activation and thus reduces the porosity. Doping drastically reduces the specific surface area, measured both by gas adsorption and small angle X-ray scattering. The latter technique demonstrates that in both doped and undoped samples the specific surface area is isotropic. X-ray photoelectron spectroscopy (XPS) reveals that the spatial distribution of Na and Si atoms within the samples are not identical. The open honeycomb structure, conserved during the heat treatment from the original wood, provides easy access for gas adsorption and separation applications. The ratios of the microporous diffusion time constants of N₂ and O₂ from frequency response (FR) measurements gave separation factors 3.0, 4.3, 2.7 and 1.3 for samples prepared at 600 °C, 700 °C, 800 °C and 900 °C, respectively.     

Dogs' ( Canis familaris ) responsiveness to human pointing gestures.

In a series of 3 experiments, dogs (Canisfamiliaris) were presented with variations of the human pointing gesture; gestures with reversed direction of movement, cross-pointing, and different arm extensions. Dogs performed at above chance level if they could see the hand (and index finger) protruding from the human body contour. If these minimum requirements were not accessible, dogs still could rely on the body position of the signaler. The direction of movement of the pointing arm did not influence the performance. In summary, these observations suggest that dogs are able to rely on relatively novel gestural forms of the human communicative pointing gesture and that they are able to comprehend to some extent the referential nature of human pointing. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

High Catalytic Activity in CO Oxidation over MnO x Nanocrystals

Manganese oxides of various stoichiometry were prepared via Mn-oxalate precipitation followed by thermal decomposition in the presence of oxygen. A non-stoichiometric manganese oxide, MnO _x (x = 1.61…1.67) was obtained by annealing at 633 K and demonstrated superior CO oxidation activity, i.e. full CO conversion at room temperature and below. The activity gradually decreased with time-on-stream of the reactants but could be easily recovered by heating at 633 K in the presence of oxygen. CO oxidation over MnO _x in the absence of oxygen proved to be possible with reduced rates and demonstrated a Mars—van Krevelen—type mechanism to be in operation. A TEM structural analysis showed the MnO _x phase to form microrods with large aspect ratio which broke up into nanocrystalline manganese oxide (MnO _x ) particles with diameters below 3 nm and a BET specific surface area of 525 m²/g. Annealing at 798 K rather than 633 K produced well crystalline Mn₂O₃ which showed lower CO oxidation activity, i.e. 100% CO conversion at 335 K. The catalytic performance in CO oxidation of various Mn-oxides either studied in this work or elsewhere was compared on the basis of specific reaction rates.     

Impact of a Stereocentre Inversion in Cyclic Lipodepsipeptides from the Viscosin Group: A Comparative Study of the Viscosinamide and Pseudodesmin Conformation and Self‐Assembly

The viscosin group covers a series of cyclic lipodepsipeptides (CLPs) produced by Pseudomonas bacteria, with a range of biological functions and antimicrobial activities. Their oligopeptide moieties are composed of both L ‐ and D ‐amino acids. Remarkably, the Leu5 amino acid—centrally located in the nonapeptide sequence—is the sole residue found to possess either an L or D configuration, depending on the producing strain. The impact of this D /L switch on the solution conformation was investigated by NMR‐restrained molecular modelling of the epimers pseudodesmin A and viscosinamide A. Although the backbone fold remained unaffected, the D /L switch adjusted the segregation between hydrophobic and hydrophilic residues, and thus the amphipathicity. It also influenced the self‐assembly capacity in organic solvents. Additionally, several new minor variants of viscosinamide A from Pseudomonas fluorescens DR54 were identified, and an NMR assay is proposed to assess the presence of either an L ‐ or D ‐Leu5.     

Surface chemistry and contrast-modified SAXS in polymer-based activated carbons

The adsorption of non-polar and polar molecules, n-hexane and water, on activated carbons, functionalized by oxidation with concentrated nitric acid, is investigated by small angle X-ray scattering. A relative mass density function p(q) is introduced in order to trace the filling characteristics of these probe molecules in the pore structure. Inspection of p(q) shows that while the pores affected by the applied relative pressure are completely filled by n-hexane, pore filling by water is only partial, even in the most oxidized carbon. The differences between the solid-liquid contact areas of these two molecules adsorbed from the vapour phase on the various carbons illustrate the influence of surface chemistry and molecular polarity on the contrast-modified SAXS response.     

Effect of transglutaminase on the quality of wheat-based pasta products

Transglutaminase (TG) enzyme was tested in T. aestivum and T. durum wheat-based model systems to evaluate the quality improvement to pasta imparted by cross-linking. Quality was measured by the enhancement in sensory and cooking properties. In this process, tests were also performed to establish the biochemical structure-functions of the proteins involved at TG enzyme concentrations between 10 and 200 mg kg⁻¹. It was observed that the amounts of water/salt-, the alcohol- and the alkali-soluble protein fractions were reduced significantly. Change in the molecular weight distribution assessed by SDS PAGE was also noticed. The sensory properties were observed to improve from high water uptake and low cooking loss. The TG treatment changed the gluten structure in the modified pasta products. Immune-reactive gliadin fractions were detected by immunoblotting and independently measured by competitive indirect ELISA, using the anti-gliadin rabbit IgG.