Boosted feature selectors: a case study on prediction P-gp inhibitors and substrates

Feature selection is commonly used as a preprocessing step to machine learning for improving learning performance, lowering computational complexity and facilitating model interpretation. This paper proposes the application of boosting feature selection to improve the classification performance of standard feature selection algorithms evaluated for the prediction of P-gp inhibitors and substrates. Two well-known classification algorithms, decision trees and support vector machines, were used to classify the chemical compounds. The experimental results showed better performance for boosting feature selection with respect to the standard feature selection algorithms while maintaining the capability for feature reduction.     

Influence of the melting conditions of heavy metal oxide glasses containing bismuth oxide on their optical absorption

Glasses of the systems Bi₂O₃-SiO₂, Bi₂O₃-PbO-Ga₂O₃, Bi₂O₃-PbO-Ga₂O₃-GeO₂ and Bi₂O₃-GeO₂-Li₂O have been prepared and the interaction of their melts with crucibles of different materials has been analytically determined. Silica and porcelain crucibles were very strongly corroded and the glass composition was noticeably altered. Instead platinum crucibles are not affected if the Bi₂O₃ content is not too high. The color of the glasses changes in all cases from pale yellow to deep brown when the melting temperature reaches approximately 1000 °C. The higher the temperature and the Bi₂O₃ content the darker the brown color, independently of the nature of the employed crucible. The addition of oxidizing ions (Sb⁵⁺, As⁵⁺ or Ce⁴⁺) to the glass batch prevents darkening. Nanoparticles of elementary bismuth Bi⁰ are observed by transmission electron microscopy in the glasses melted above 1000 °C. The partial thermoreduction of the Bi₂O₃ during the heating of the glass melt is proposed as the mechanism responsible for the observed darkening.     

Anthracene-Fused Oligo-BODIPYs: A New Class of π-Extended NIR-Absorbing Materials

Large π-conjugated systems are key in the area of molecular materials. Herein, we prepare via Au^I-catalyzed cyclization a series of fully π-conjugated anthracene-fused oligo-BODIPYs. Their structural and optoelectronic properties were studied by several techniques, ranging from X-ray, UV/Vis, and cyclic voltammetry to transient absorption spectroscopy. As a complement, their electronic structures were explored by means of Density Functional Theory (DFT) calculations. Depending on the size and shape of the π-conjugated skeleton, unique features—such as face-to-face supramolecular organization, NIR absorption and fluorescence as well as strong electron accepting character—were noted. All in all, the aforementioned features render them valuable for technological applications.     

Merging the Isonitrile-Tetrazine (4+1) Cycloaddition and the Ugi Four-Component Reaction into a Single Multicomponent Process

Multicomponent reactions are of utmost importance at generating a unique, wide, and complex chemical space. Herein we describe a novel multicomponent approach based on the combination of the isonitrile-tetrazine (4+1) cycloaddition and the Ugi four-component reaction to generate pyrazole amide derivatives. The scope of the reaction as well as mechanistic insights governing the 4H-pyrazol-4-imine tautomerization are provided. This multicomponent process provides access to a new chemical space of pyrazole amide derivatives and offers a tool for peptide modification and stapling.     

Macro‐hybrid mixed variational two‐phase flow through fractured porous media

Macro‐hybrid mixed variational models of two‐phase flow, through fractured porous media, are analyzed at the mesoscopic and macroscopic levels. The mesoscopic models are treated in terms of nonoverlapping domain decompositions, in such a manner that the porous rock matrix system and the fracture network interact across rock–rock, rock–fracture, and fracture–fracture interfaces, with flux transmission conditions dualized. Alternatively, the models are scaled to a macroscopic level via an asymptotic process, where the width of the fractures tends to zero, and the fracture network turns out to be an interface system of one less spatial dimension, with variable high permeability. The two‐phase flow is characterized by a fractional flow dual mixed variational model. Augmented two‐field and three‐field variational reformulations are presented for regularization, internal approximations, and macro‐hybrid mixed finite element implementation. Also abstract proximal‐point penalty‐duality algorithms are derived and analyzed for parallel computing. Copyright © 2016 John Wiley & Sons, Ltd.     

On the Ramsey number of the triangle and the cube

The Ramsey number r(K ₃,Q _n ) is the smallest integer N such that every red-blue colouring of the edges of the complete graph K _N contains either a red n-dimensional hypercube, or a blue triangle. Almost thirty years ago, Burr and Erd?s conjectured that r(K ₃,Q _n )=2 ⁿ⁺¹?1 for every n∈?, but the first non-trivial upper bound was obtained only recently, by Conlon, Fox, Lee and Sudakov, who proved that r(K ₃,Q _n )?7000·2 ⁿ . Here we show that r(K ₃,Q _n )=(1+o(1))2 ⁿ⁺¹ as n→∞.     

Organocatalytic Enantioselective Synthesis of Pyrazoles Bearing a Quaternary Stereocenter

An efficient one‐pot asymmetric synthesis of pyrazoles bearing a chiral quaternary stereocenter has been developed. Quinine‐derived thiourea catalyzed the enantioselective addition of pyrazolones to isatin‐derived ketimines, providing the corresponding acetylated pyrazoles after in situ treatment with Ac₂O/Et₃N. The corresponding pyrazoles were afforded in high yields and excellent enantioselectivities.     

Inductive vs solvation effects in primary alkyl amines: determination of the standard potentials

The determination of the standard potential of alkyl primary amines is reported for the first time using the nanosecond equilibrium method. The versatility and accuracy of the method demonstrates that it is not only an alternative to the classical and modern electrochemical methods, but also a powerful tool for quantifying inductive and/or solvation effects in a related family of compounds. Two different trends were observed depending on alkyl chain length. For "short-chain" alkyl primary amines, where the solvation around the amino group is expected to be the same, the standard potential value appears to follow a linear relationship with the number of carbon atoms, which indicates that the methylene group (-CH2-) causes an inductive effect that is responsible for the stabilization of the amine cation radical. Meanwhile, the E(o) rises slightly to a constant potential value 1.500 V for "long-chain" unbranched alkyl primary amines. This interesting result can be explained by a steric inhibition of solvation around the amino group due to a fold of the long alkyl chain following a solvent exclusion mechanism.