Simulations of full multivariate Tweedie with flexible dependence structure

We employ a variables-in-common method for constructing multivariate Tweedie distributions, based on linear combinations of independent univariate Tweedie variables. The method lies on the convolution and scaling properties of the Tweedie laws, using the cumulant generating function for characterization of the distributions and correlation structure. The routine allows the equivalence between independence and zero correlation and gives a parametrization through given values of the mean vector and dispersion matrix, similarly to the Gaussian vector. Our approach leads to a matrix representation of multivariate Tweedie models, which permits the simulations of many known distributions, including Gaussian, Poisson, non-central gamma, gamma, and inverse Gaussian, both positively or negatively correlated.     

Drying of guaco leaves: experimental and modeling kinetic,equilibrium isotherms and heat of desorption

Journal of Thermal Analysis and Calorimetry - Due to the huge variety of therapeutic properties found in plant leaves, many of these are widely used in traditional medicine. Among these plants,...     

Batch tautomer generation with MolTPC

Besides all their conformational degrees of freedom, drug‐like molecules and natural products often also undergo tautomeric interconversions. Compared to the huge efforts made in experimental investigation of tautomerism, open and free algorithmic solutions for prototropic tautomer generation are surprisingly rare. The few freely available software packages limit their output to a subset of the possible configurational space by sometimes unwanted prior assumptions and complete neglection of ring‐chain tautomerism. Here, we describe an adjustable fully automatic tautomer enumeration approach, which is freely available and also incorporates the detection of ring‐chain variants. The algorithm is implemented in the MolTPC framework and accessible on SourceForge. © 2013 Wiley Periodicals, Inc.     

Mechanistic Insights into the Rhenium‐Catalyzed Alcohol‐To‐Olefin Dehydration Reaction

Rhenium‐based complexes are powerful catalysts for the dehydration of various alcohols to the corresponding olefins. Here, we report on both experimental and theoretical (DFT) studies into the mechanism of the rhenium‐catalyzed dehydration of alcohols to olefins in general, and the methyltrioxorhenium‐catalyzed dehydration of 1‐phenylethanol to styrene in particular. The experimental and theoretical studies are in good agreement, both showing the involvement of several proton transfers, and of a carbenium ion intermediate in the catalytic cycle.     

Recent Progress in the Simulation of Chiral Systems with Real Time Propagation Methods

In this brief review, an overview about recent efforts to simulate the spectroscopic signatures of chiral molecules is given with focus on real time propagation approaches to solve the time-dependent Schrödinger equation. In particular the simulation of electric circular dichroism spectra and vibrational Raman optical activity is discussed. In comparison to linear absorption spectra, where only the response of the electric dipole moment is necessary, the response of the magnetic dipole moment and electric quadrupole moment is more intricate. Issues such as gauge origin dependence, basis set dependence, non-local potentials and the dipole approximation are addressed.     

ChemStable: a web server for rule-embedded naïve Bayesian learning approach to predict compound stability

Predicting compound chemical stability is important because unstable compounds can lead to either false positive or to false negative conclusions in bioassays. Experimental data (COMDECOM) measured from DMSO/H₂O solutions stored at 50 °C for 105 days were used to predicted stability by applying rule-embedded naïve Bayesian learning, based upon atom center fragment (ACF) features. To build the naïve Bayesian classifier, we derived ACF features from 9,746 compounds in the COMDECOM dataset. By recursively applying naïve Bayesian learning from the data set, each ACF is assigned with an expected stable probability (p _s ) and an unstable probability (p _uns ). 13,340 ACFs, together with their p _s and p _uns data, were stored in a knowledge base for use by the Bayesian classifier. For a given compound, its ACFs were derived from its structure connection table with the same protocol used to drive ACFs from the training data. Then, the Bayesian classifier assigned p _s and p _uns values to the compound ACFs by a structural pattern recognition algorithm, which was implemented in-house. Compound instability is calculated, with Bayes’ theorem, based upon the p _s and p _uns values of the compound ACFs. We were able to achieve performance with an AUC value of 84 % and a tenfold cross validation accuracy of 76.5 %. To reduce false negatives, a rule-based approach has been embedded in the classifier. The rule-based module allows the program to improve its predictivity by expanding its compound instability knowledge base, thus further reducing the possibility of false negatives. To our knowledge, this is the first in silico prediction service for the prediction of the stabilities of organic compounds.     

Tapered copolymers of styrene and 4-vinylbenzocyclobutene via carbanionic polymerization for crosslinkable polymer films

Well-defined polystyrene homopolymers with surface-adhesive triethoxysilyl end group were synthesized via living carbanionic polymerization, epoxide end-functionalization and subsequent hydrosilylation with triethoxysilane. Grafting-to performance of polymers with various molecular weight (M_n = 3000–14,000 g mol⁻¹) to a silicon surface was examined in dependence of reaction time, polymer concentration, solvent and number of alkoxysilyl end groups. Crosslinkable polymers for surface modification were synthesized by statistical carbanionic copolymerization of 4-vinylbenzocyclobutene (4-VBCB) and styrene, followed by epoxide end-functionalization and triethoxysilane modification (M_n = 4000–14,000 g mol⁻¹). The copolymers were characterized by ¹H-NMR, THF-SEC, and matrix-assisted laser desorption and ionization time-of-flight mass spectrometry. In situ ¹H-NMR kinetic studies in cyclohexane-d₁₂ provided information regarding the monomer gradient in the polymer chains, with styrene being the more reactive monomer (r_s = 2.75, r_4-VBCB = 0.23). Thin polymer films on silicon wafers were prepared by grafting-to surface modification under conditions derived for the polystyrene homopolymer. The traceless, thermally induced crosslinking reaction of the benzocyclobutene units was studied by DSC in bulk as well as in 3–6 nm thick polymer films. Crosslinked films were analyzed by atomic force microscopy, ellipsometry, and nanoindentation, showing smooth polymer films with an increased modulus. © 2019 The Authors. Journal of Polymer Science published by Wiley Periodicals, Inc. J. Polym. Sci. 2020 , 58, 181–192     

An Exploratory Study on the Complexity and Machine Learning Predictability of Stock Market Data

This paper shows if and how the predictability and complexity of stock market data changed over the last half-century and what influence the M1 money supply has. We use three different machine learning algorithms, i.e., a stochastic gradient descent linear regression, a lasso regression, and an XGBoost tree regression, to test the predictability of two stock market indices, the Dow Jones Industrial Average and the NASDAQ (National Association of Securities Dealers Automated Quotations) Composite. In addition, all data under study are discussed in the context of a variety of measures of signal complexity. The results of this complexity analysis are then linked with the machine learning results to discover trends and correlations between predictability and complexity. Our results show a decrease in predictability and an increase in complexity for more recent years. We find a correlation between approximate entropy, sample entropy, and the predictability of the employed machine learning algorithms on the data under study. This link between the predictability of machine learning algorithms and the mentioned entropy measures has not been shown before. It should be considered when analyzing and predicting complex time series data, e.g., stock market data, to e.g., identify regions of increased predictability.