An exploratory study using QICAR models for prediction of adsorption capacity of multi-walled carbon nanotubes for heavy metal ions.

The Quantitative Ion Character-Activity Relationship (QICAR) method was used for correlating metal ionic characteristics with the maximum adsorption capacity (qmax) of multi-walled carbon for heavy metals. The experimental values of qmax for 25 heavy metal ions, estimated by the Langmuir isotherm model, were used to construct a QICAR model. The genetic algorithm, enhanced replacement method and successive projection algorithm procedures were applied as variable selection algorithms to choose the optimal subsets of descriptors. The selected variables were correlated with qmax values by using partial least squares (PLS) regression. Orthogonal signal correction was applied as a pre-processing technique. Among of different variable selection methods, the enhanced replacement method displayed noticeable statistical parameters of the final model. The results of the enhancement replacement method-orthogonal correction signal-PLS model, with RMSEC = 0.733, r2c = 0.999 and r2p = 0.946, were excellent and dramatically better than those of other models. The developed QICAR model satisfied the internal and external validation criteria. The importance of electronegativity, ionic radius and atomic number of the heavy metal ions indicated the impact of the tendency to accept electrons and the size of ions in adsorption on carbon nanotubes.

Global and local QSPR models to predict supercooled vapour pressure for organic compounds.

In this study, a quantitative structure-property relationship (QSPR) approach was used for estimation of logarithmic values of supercooled liquid vapour pressure (log PL) of a large set of structurally diverse organic compounds. This set includes 12 local sets of aromatic and aliphatic hydrocarbons, polychlorinated biphenyls, ethers, polychlorinated and brominated diphenylethers, polychlorinated naphthalenes and alcohols. Some simple models based on the linear relationship between log PL and VolSurf descriptors were developed as global models, and a general equation as a simple way to calculate the supercooled liquid vapour pressure of organic chemicals was provided. A descriptor representing the hydrophilic regions (WO1) of organic chemicals showed the highest correlation with log PL and resulted in a one-parameter global model characterized by satisfactory statistical performance; calibration (r2c) and prediction (r2p) correlation coefficient of 0.84 and 0.85, respectively. Moreover, local QSPR models were also developed for each subset of organic compounds and, as expected, the statistical results obtained from these models were better than the global one. From the descriptors involved in the models, it is concluded that the hydrophilic and hydrophobic regions at different energy levels and polarizability usually determine the variation of supercooled liquid vapour pressure of organic compounds.

3D-QSAR studies on the toxicity of substituted benzenes to Tetrahymena pyriformis: CoMFA, CoMSIA and VolSurf approaches.

Three-dimensional quantitative structure-activity relationship (3D-QSAR) analysis were performed on the toxicity of a large set of substituted benzenes toward ciliate Tetrahymena pyriformis. The 3D-QSAR studies were carried out using comparative molecular field analysis (CoMFA), comparative molecular similarity indices analysis (CoMSIA) and VolSurf techniques. The optimal CoMFA and CoMSIA models obtained from the training set were all statistically significant with correlation coefficients (R(2)) greater than 0.79 and absolute error less than 0.33 in log units. The predictive ability of the models was externally evaluated through the prediction of a test set (20 percent of the whole data set) that were not included in the training set. A simple and fairly good predictive linear model based on VolSurf descriptors was also developed that showed an adequate prediction power of the toxicity (pIGC50) of substituted benzenes. Validation, reliability and robustness of models were also evaluated by leave-one-out, leave-four-out, bootstrapping and progressive scrambling approaches. The results confirmed that in addition to hydrophobic effects, electrostatic and H-bonding interactions also play important roles in the toxicity of substituted benzenes. The information obtained from CoMFA and CoMSIA 3-D contour maps could be useful to explain the toxicity mechanism of substituted benzenes.

Comparative molecular field analysis and comparative molecular similarity index analysis studies on 1H NMR chemical shift of NH group of diaryl triazene derivatives.

Comparative molecular field analysis (CoMFA), comparative molecular field analysis region focusing (CoMFA-RF) for optimizing the region for the final partial least square analysis, and comparative molecular similarity indices analysis (CoMSIA) methods were employed to develop three-dimensional quantitative structure-activity relationship (3D-QSAR) models of (1)H NMR chemical shift of NH proton of diaryl triazene derivatives. The best orientation was searched by all-orientation search (AOS) strategy to minimize the effect of the initial orientation of the structures. The predictive abilities of CoMFA-RF and CoMSIA models were determined using a test set of ten compounds affording predictive correlation coefficients of 0.721 and 0.754, respectively, indicating good predictive power. For further model validation, cross validation (leave one out), progressive scrambling, and bootstrapping were also applied. The accuracy and speed of obtained 3D-QSAR models for the prediction of (1)H NMR chemical shifts of NH group of diaryl triazene derivatives were greater compared to some computational well-known procedures.