QSAR modelling using combined simple competitive learning networks and RBF neural networks

The aim of this study was to propose a QSAR modelling approach based on the combination of simple competitive learning (SCL) networks with radial basis function (RBF) neural networks for predicting the biological activity of chemical compounds. The proposed QSAR method consisted of two phases. In the first phase, an SCL network was applied to determine the centres of an RBF neural network. In the second phase, the RBF neural network was used to predict the biological activity of various phenols and Rho kinase (ROCK) inhibitors. The predictive ability of the proposed QSAR models was evaluated and compared with other QSAR models using external validation. The results of this study showed that the proposed QSAR modelling approach leads to better performances than other models in predicting the biological activity of chemical compounds. This indicated the efficiency of simple competitive learning networks in determining the centres of RBF neural networks.     

Comparative study to predict toxic modes of action of phenols from molecular structures

Quantitative structure–activity relationship models for the prediction of mode of toxic action (MOA) of 221 phenols to the ciliated protozoan Tetrahymena pyriformis using atom-based quadratic indices are reported. The phenols represent a variety of MOAs including polar narcotics, weak acid respiratory uncouplers, pro-electrophiles and soft electrophiles. Linear discriminant analysis (LDA), and four machine learning techniques (ML), namely k-nearest neighbours (k-NN), support vector machine (SVM), classification trees (CTs) and artificial neural networks (ANNs), have been used to develop several models with higher accuracies and predictive capabilities for distinguishing between four MOAs. Most of them showed global accuracy of over 90%, and false alarm rate values were below 2.9% for the training set. Cross-validation, complementary subsets and external test set were performed, with good behaviour in all cases. Our models compare favourably with other previously published models, and in general the models obtained with ML techniques show better results than those developed with linear techniques. We developed unsupervised and supervised consensus, and these results were better than our ML models, the results of rule-based approach and other ensemble models previously published. This investigation highlights the merits of ML-based techniques as an alternative to other more traditional methods for modelling MOA.     

Model polyurethane elastomers prepared from noncrystallizable poly(propylene oxide) chains

Elastomers of controlled molecular structure were prepared from hydroxyl-terminated atactic poly(propylene oxide) (PPO) chains having number-average molecular weights M_n in the range 800–4360 g mole^?1. The chains were end-linked into noncrystallizable trifunctional networks using a specially prepared aromatic triisocyanate. The networks thus obtained were studied with regard to their stress–strain isotherms in the unswollen state, in elongation at 25°C, and with regard to their equilibrium swelling in benzene at 61°C. Values of the modulus in the limit at high deformation were in good agreement with corresponding results previously obtained on networks of poly(dimethylsiloxane) (PDMS). This is of considerable importance since use of the widely used “plateau modulus” as a measure of interchain entangling would suggest that the networks of PPO would have a much higher density of such entanglements than would the corresponding networks of PDMS. The close similarity between the moduli of the two types of networks therefore argues against the idea that such entanglements make large contributions to the equilibrium elastomeric properties of a polymer network. These values of the high deformation modulus are also in good agreement with recent molecular theories as applied to the nonaffine deformation of a “phantom” network. The values of the low deformation modulus were considerably smaller than the values predicted for an affine deformation, however, suggesting that the junction points were not firmly embedded in the network structure. This is presumably due to the relatively low degree of chain-junction entangling in the case of relatively short network chains. The swelling equilibrium results were in very good agreement with the new theory of network swelling developed by Flory.     

The Learned Symmetry Concept in Revealing Quantitative Structure-Activity Relationships With Artificial Neural Networks

Abstract

A method to build QSAR models based on substituent constants for congeneric sets of compounds having several topologically equivalent substituent positions was proposed. The approach is based on the application of artificial neural networks (learning to construct nonlinear structure-activity relationships taking into account necessary symmetry properties of training set structures) to a training set expanded by adding the copies of compounds with the same activity values but with permuted assignment of equivalent substituent positions. The better predictive power of these constructed models, as compared with the performances of neural network models for non-expanded sets was demonstrated for the calcium channel blockers of 1,4-dihydropyridine type and for hallucinogenic phenylalkylamines.     

Identification of rhubarbs by using NIR spectrometry and temperature-constrained cascade correlation networks

Temperature-constrained cascade correlation networks (TCCCNs) were used to identify powdered rhubarbs based on their near-infrared spectra. Different network configurations that used multiple network models with single output (Uni-TCCCN) and single networks with multiple outputs (Multi-TCCCN) were compared. Comparative studies were made by using Latin-partitions and leave-one-out cross-validation methods. Results showed that multiple networks with single output predicted generally better than single network with multiple outputs. Better results with TCCCN models were obtained compared with conventional back propagation neural networks (BPNNs). The effects of parameters on correct identification and parameter optimizations were discussed in detail. With optimized neural network training parameters, NIR spectra from powdered rhubarb samples were classified by a TCCCN model with 100% accuracy.     

A Quantitative Structure–Activity Relationship Study of Calpeptin (Calpain Inhibitor) as an Anticancer Agent

Calpeptin analogs show anticancer properties with inhibition of calpain. In this work, we applied a quantitative structure–activity relationship (QSAR) model on 34 calpeptin derivatives to select the most appropriate compound. QSAR was employed to generate the models and predict the more significant compounds through a series of calpeptin derivatives. The HyperChem, Gaussian 09, and Dragon software programs were used for geometry optimization of the molecules. The 2D and 3D molecular structures were drawn by ChemDraw (Ultra 16.0) and Chem3D (Pro16.0) software. The Unscrambler program was used for the analysis of data. Multiple linear regression (MLR‐MLR), partial least‐squares (MLR‐PLS1), principal component regression (MLR‐PCR), a genetic algorithm‐artificial neural networks (GA‐ANN), and a novel similarity analysis‐artificial neural network (SA‐ANN) method were used to create QSAR models. Among the three MLR models, MLR‐MLR provided better statistical parameters. The R² and RMSE of the prediction were estimated as 0.8248 and 0.26, respectively. Nevertheless, the constructed model using GA‐ANN revealed the best statistical parameters among the studied methods (R² test = 0.9643, RMSE test = 0.0155, R² train = 0.9644, RMSE train = 0.0139). The GA‐ANN model is found to be the most favorable method among the statistical methods and can be employed for designing new calpeptin analogs as potent calpain inhibitors in cancer treatment.     

Application of artificial neural network to simultaneous potentiometric determination of silver(I), mercury(II) and copper(II) ions by an unmodified carbon paste electrode

The response characteristics and selectivity coefficients of an unmodified carbon paste electrode (CPEs) towards Ag⁺, Cu²⁺ and Hg²⁺ were evaluated. The electrode was used as an indicator electrode for the simultaneous determination of the three metal ions in their mixtures via potentiometric titration with a standard thiocyanate solution. A three-layered feed-forward artificial neural network (ANN) trained by back-propagation learning algorithm was used to model the complex non-linear relationship between the concentration of silver, copper and mercury in their different mixtures and the potential of solution at different volumes of the added titrant. The network architecture and parameters were optimized to give low prediction errors. The optimized networks were able to precisely predict the concentrations of the three cations in synthetic mixtures.     

Synthesis and characterization of an environmentally friendly PHBV/PEG copolymer network as a phase change material

Novel environmentally friendly poly(hydroxybutyrate-co-hydroxyvalerate) and poly(ethylene glycol) (PHBV/PEG) copolymer networks were synthesized through free-radical solution polymerization with PHBV diacrylate (PHBVDA) and polyethylene glycol diacrylate (PEGDA) as macromers. The molecular structure of PHBV/PEG copolymer network was characterized by Fourier transform infrared (FT-IR) and ¹H nuclear magnetic resonance (¹H NMR). The morphology of the PHBV/PEG copolymer network was characterized by polarization optical microscopy. Thermal energy storage properties, thermal reliability and thermal stability were investigated by differential scanning calorimetry (DSC) and thermogravimetric analysis. The results indicated that the PHBV/PEG copolymer network hindered the growth of PEG crystalline segments or PHBV segments. PHBV/PEG copolymer network had a higher latent heat enthalpy, which didn’t reduce with the components of PHBV increased. Moreover, PHBV/PEG copolymer network still had good thermal stability even at 300 °C. These results suggested that such environmentally friendly copolymer network would have wide applications in phase change energy storage materials.     

人工神经网络用于预测离子交换分配系数

用前向神经和扩展的ｄｅｌｔａ－ｂａｒ－ｄｅｌｔａ算法对主族特征价阳离子的离子交换分配系数（Ｋｄ）进行了预测，对网络结构，学习次数进行了优化并研究了学习集的大小，１ｎＫｄ的均方根偏差的（ＲＭＳ）小于７％。     

Structure and properties of cellulose/poly(N‐isopropylacrylamide) hydrogels prepared by IPN strategy

Interpenetrating polymer network (IPN) strategy was developed to fabricate novel hydrogels composed of cellulose and poly(N‐isopropylacrylamide) (PNIPAAm) with high mechanical strength and adjustable thermosensitivity. Cellulose hydrogels were prepared by chemically cross‐linking cellulose in NaOH/urea aqueous solution, which were employed as the first network. The second network was subsequently obtained by in situ polymerization/cross‐linking of N‐isopropylacrylamide in the cellulose hydrogels. The results from FTIR and solid ¹³C NMR indicated that the two networks co‐existed in the IPN hydrogels, which exhibited uniform porous structure, as a result of good compatibility. The mechanical and swelling properties of IPN hydrogels were strongly dependent on the weight ratio of two networks. Their temperature‐sensitive behaviors and deswelling kinetics were also discussed. This work created double network hydrogels, which combined the advantages of natural polymer and synthesized PNIPAAm collectively in one system, leading to the controllable temperature response and improvement in the physical properties. Copyright © 2009 John Wiley & Sons, Ltd.     

Qsar Models for Predicting the Acute Toxicity of Selected Organic Chemicals with Diverse Structures to Aquatic Non-Vertebrates and Humans

Abstract

The linear and non-linear relationships of acute toxicity (as determined on five aquatic non-vertebrates and humans) to molecular structure have been investigated on 38 structurally-diverse chemicals. The compounds selected are the organic chemicals from the 50 priority chemicals prescribed by the Multicentre Evaluation of In Vitro Cytotoxicity (MEIC) programme. The models used for the evaluations are the best combination of physico-chemical properties that could be obtained so far for each organism, using the partial least squares projection to latent structures (PLS) regression method and backpropagated neural networks (BPN). Non-linear models, whether derived from PLS regression or backpropagated neural networks, appear to be better than linear models for describing the relationship between acute toxicity and molecular structure. BPN models, in turn, outperform non-linear models obtained from PLS regression. The predictive power of BPN models for the crustacean test species are better than the model for humans (based on human lethal concentration). The physico-chemical properties found to be important to predict both human acute toxicity and the toxicity to aquatic non-vertebrates are the n?octanol water partition coefficient (Pow) and heat of formation (HF). Aside from the two former properties, the contribution of parameters that reflect size and electronic properties of the molecule to the model is also high, but the type of physico-chemical properties differs from one model to another. In all of the best BPN models, some of the principal component analysis (PCA) scores of the ¹³C-NMR spectrum, with electron withdrawing/accepting capacity (LUMO, HOMO and IP) are molecular size/volume (VDW or MS1) parameters are relevant. The chemical deviating from the QSAR models include non-pesticides as well as some of the pesticides tested. The latter type of chemical fits in a number of the QSAR models. Outliers for one species may be different from those of other test organisms.     

Composition,structural characteristics,and antitumor properties of polysaccharides from the brown algae <Emphasis Type="Italic">Dictyopteris polypodioides</Emphasis> and <Emphasis Type="Italic">Sargassum</Emphasis> sp.

The polysaccharide compositions of the brown algae Dictyopteris polypodioides and Sargassum sp. from the Mediterranean Sea were determined. The principal polysaccharide of the studied algae (about 12% of the dry alga weight) was alginic acid. The content of water-soluble polysaccharides was low. The amount of fucoidan was less than 1% of the dry alga weight; of neutral polysaccharides, less than 0.25%. The monosaccharide compositions of fucoidans and neutral polysaccharides were investigated. Experiments on soft agar-agar models showed that fucoidans from D. polypodioides and Sargassum sp. exhibited antitumor activity against RPMI-7951 human melanoma cells.     

Concept of hard clusters in the interpretation of thermal and mechanical properties of polyurethane and polyurethane acrylate networks

Polyurethane (PU) and polyurethane acrylate (PUA) networks based on hydroxyl-terminated polycaprolactone (PCL), 1,3-bis-2,2′(2-isocyanatopropyl)benzene (m-TMXDI), trimethylolpropane (TMP) for PU or hydroxyethyl methacrylate (HEMA) for PUA were synthesized. Glass transition temperature, T_g, dynamic mechanical relaxation, α, and equilibrium tensile modulus, E′, were measured to compare the two kinds of networks. To explain thermal and mechanical properties of networks, the concept of hard clusters has been introduced. PU networks exhibit a single-phase structure with modulus and T_g dependent on the concentration of elastically active network chains (EANC) per unit volume calculated by considering hard crosslink clusters. The rigidity of the clusters comes from small diisocyanate and trimethylolpropane units connected by urethane bonds. They are embedded in a continuous soft phase of macrodiol urethane. Physical equivalence between several kinds of network models has been demonstrated for full conversion of isocyanate-alcohol reaction. PUA networks exhibit thermodynamically one-phase structures that become a two-phase structure for high molar mass of macrodiol when the molar fraction of isocyanate groups increases. For those networks, the calculated modulus considering clusters based on polyacrylate chains seems to be a good way to approach the experimental value of the equilibrium modulus. For the same molar ratio of OH to NCO groups the range of dynamic moduli is larger for PUA than for PU. This difference can be explained by a different concentration of crosslinks in the networks. © 1996 John Wiley & Sons, Inc.     

Quality control of the powder pharmaceutical samples of sulfaguanidine by using NIR reflectance spectrometry and temperature-constrained cascade correlation networks

Temperature-constrained cascade correlation networks (TCCCNs) were applied to the identification of the powder pharmaceutical samples of sulfaguanidine based on near infrared (NIR) diffuse reflectance spectra and their first derivative spectra. This work focused on the comparison of performances of the uni-output TCCCN (Uni-TCCCN) and multi-output (Multi-TCCCN) by near infrared diffuse reflectance spectra and their first derivative spectra of sulfaguanidine. The TCCCN models were verified with independent prediction samples by using the “cross-validation” method. The networks were used to discriminate qualified, un-qualified and counterfeit sulfaguanidines pharmaceutical powders. The results showed that single outputs network generally worked better than the multiple outputs networks, and the first derivative spectra were more suitable for the identification comparing with original diffuse reflectance spectra. With proper network parameters the pharmaceutical powders can be classified at rate of 100% in this work. Also, the effects of parameters and related problems were discussed.     

Simultaneous kinetic determination of sulfite and sulfide using artificial neural networks

Artificial neural networks (ANNs) are proposed for the determination of sulfite and sulfide simultaneously. The method is based on the reaction between Brilliant Green (BG) as a colored reagent and sulfite and/or sulfide in buffered solution (pH 7.0) and monitoring the changes of absorbance at maximum wavelength of 628 nm. Experimental conditions such as pH, reagents concentrations, and temperature were optimized and training the network was performed using principal components (PCs) of the original data. The network architecture (number of input, hidden and output nodes), and some parameters such as learning rate (η) and momentum (α) were also optimized for getting satisfactory results with minimum errors. The measuring range was 0.05-3.6 μg ml⁻¹ for both analytes. The proposed method has been successfully applied to the quantification of the sulfite and sulfide in different water samples.     

Identification of combustible material with piezoelectric crystal sensor array using pattern-recognition techniques

A promising way of increasing the selectivity and sensitivity of gas sensors is to treat the signals from a number of different gas sensors with pattern recognition (PR) method. A gas sensor array with seven piezoelectric crystals each coated with a different partially selective coating material was constructed to identify four kinds of combustible materials which generate smoke containing different components. The signals from the sensors were analyzed with both conventional multivariate analysis, stepwise discriminant analysis (SDA), and artificial neural networks (ANN) models. The results show that the predictions were even better with ANN models. In our experiment, we have reported a new method for training data selection, 'training set stepwise expending method' to solve the problem that the network can not converge at the beginning of the training. We also discussed how the parameters of neural networks, learning rate eta, momentum term alpha and few bad training data affect the performance of neural networks.