Genetic Algorithm Variable Subset Selection Research Articles

Chemoinformatic modelling of the antioxidant activity of phenolic compounds.

Antioxidants are chemicals used to protect foods from deterioration by neutralizing free radicals and inhibiting the oxidative process. One approach to investigate the antioxidant activity is to develop quantitative structure-activity relationships (QSARs). A curated database of 165 structurally heterogenous phenolic compounds with the Trolox equivalent antioxidant capacity (TEAC) was developed. Molecular geometries were optimized by means of the GFN2-xTB semiempirical method and diverse molecular descriptors were obtained afterwards. For model development, V-WSP unsupervised variable reduction was used before performing the genetic algorithms-variable subset selection (GAs-VSS) to construct the best five-descriptor multiple linear regression model. The coefficient of determination and the root-mean-square error were used to measure the performance in calibration (R2 = 0.789 and RMSEC = 0.381), and test set prediction (Q2 = 0.748 and RMSEP = 0.416), along several cross-validation criteria. To thoroughly understand the TEAC prediction, a fully explained mechanism of action of the descriptors is provided. In addition, the applicability domain of the model defined a theoretical chemical space for reliable predictions of new phenolic compounds. This in silico model conforms to the five principles stated by the Organization for Economic Co-operation and Development (OECD). The model may be useful for virtual screening of the antioxidant chemical space, and for identifying the most potent molecules related to an experimental measurement of TEAC activity. In addition, the model could assist chemists working on computer-aided drug design (CADD) for the synthesis of new targets with improved activity and potential uses in food science. This article is protected by copyright. All rights reserved.

Modeling and prediction of octanol/water partition coefficient of pesticides using QSPR methods

PurposeThe purpose of this paper is to predict the octanol/water partition coefficient (Kow) of 43 organophosphorous compounds.Design/methodology/approachA quantitative structure-property relationship analysis was performed on a series of 43 pesticides using multiple linear regression and support vector machines methods, which correlate the octanol-water partition coefficient (Kow) values of these chemicals to their structural descriptors. At first, the data set was randomly separated into a training set (34 chemicals) and a test set (nine chemicals) for statistical external validation.FindingsModels with three descriptors were developed using theoretical descriptors as independent variables derived from Dragon software while applying genetic algorithm-variable subset selection procedure.Originality/valueThe robustness and the predictive performance of the proposed linear model were verified using both internal and external statistical validation. One influential point which reinforces the model and an outlier were highlighted.

QSPR study on solubility of some fullerenes derivatives using the genetic algorithms — Multiple linear regression

A quantitative structure–property relation study was performed on the solubility of C60 and C70 fullerene derivatives. Topological and geometrical as well as quantum mechanical energy-related and charge distribution-related descriptors, generated from CODESSA, were calculated to define the molecule structures requirement for measuring their correlations with solubility. The best four variables among the other subsets were selected by the genetic algorithm variable subset selection procedure. Modeling of the relationship between selected molecular descriptors and solubility data was achieved by multiple linear regression method (R2train=0.801, R2test=0.792, Q2LOO=0.716, Q2BOOT=0.674). The robustness and the predictive performance of the proposed model were verified using both internal (cross-validation by leave one out, bootstrap, Y-scrambling) and external statistical validations (external test set by splitting the original data set into training and test sets by k-nearest neighbor (kNN) classification method). Further, the external predictive power of the developed model was examined by considering modified r2 and concordance correlation coefficient values. The reactivity, the polar interactions, the electron–electron repulsion energy, the electronuclear attraction energy, the nuclear–nuclear repulsion energy, and the rotational–vibrational energies were the main independent factors contributing to the solubility of the fullerenes.

Estudio de la Relación Cuantitativa Estructura-Actividad de pesticidas mediante técnicas de clasificación

El objetivo de este trabajo fue la comparación entre los métodos de clasificación del vecino más cercano (κ-NN) y las redes neuronales artificiales de contrapropagación (CP-ANN) para modelar la toxicidad de un conjunto de 192 pesticidas organoclorados, organofosforados, carbamatos y piretroides, medidos como Concentración Efectiva (EC50) y que fueron divididos en tres clases, es decir, baja, intermedia y alta toxicidad. Se calcularon 4885 descriptores moleculares usando el programa DRAGON, los que fueron simultáneamente analizados mediante el método κ-NN acoplado con la técnica de selección de variables de los Algoritmos Genéticos (GA-VSS). Los modelos fueron apropiadamente validados mediante un subconjunto de predicción. Los resultados claramente sugieren que los descriptores 3D no ofrecen información relevante para modelar las clases. Por otro lado, κ-NN muestra mejores resultados que CP-ANN.

Integrated QSPR models to predict the soil sorption coefficient for a large diverse set of compounds by using different modeling methods

The soil sorption coefficient (Koc) is a key physicochemical parameter to assess the environmental risk of organic compounds. To predict soil sorption coefficient in a more effective and economical way, here, quantitative structure-property relationship (QSPR) models were developed based on a large diverse dataset including 964 non-ionic organic compounds. Multiple linear regression (MLR), local lazy regression (LLR) and least squares support vector machine (LS-SVM) were utilized to develop QSPR models based on the four most relevant theoretical molecular descriptors selected by genetic algorithms-variable subset selection (GA-VSS) procedure. The QSPR development strictly followed the OECD principles for QSPR model validation, thus great attentions were paid to internal and external validations, applicability domain and mechanistic interpretation. The obtained results indicate that the LS-SVM model performed better than the MLR and the LLR models. For best LS-SVM model, the correlation coefficients (R2) for the training set was 0.913 and concordance correlation coefficient (CCC) for the prediction set was 0.917. The root-mean square errors (RMSE) were 0.330 and 0.426, respectively. The results of internal and external validations together with applicability domain analysis indicate that the QSPR models proposed in our work are predictive and could provide a useful tool for prediction soil sorption coefficient of new compounds.

Externally validated QSPR modelling of VOC tropospheric oxidation by NO3 radicals†

The troposphere is the principal recipient of volatile organic chemicals (VOCs) of both anthropogenic and biogenic origin. The persistence of these compounds in the troposphere is an important factor for the evaluation of their fate, and the possible negative effects to the environment and human health. In this study, the tropospheric lifetime of 166 VOCs, in terms of night-time degradation rates with nitrate radical (NO3), was modelled by the quantitative structure-property relationships (QSPR) approach. The multiple linear regression method was applied, in combination with the genetic algorithm-variable subset selection procedure, to a variety of theoretical molecular descriptors, calculated by the DRAGON software. The models were developed according to the OECD principles for regulatory acceptance of QSARs, with particular attention to external validation and applicability domain (AD). The external validation was performed on an unbiased external test set or by splitting the available data using self-organized maps or the random by response approach. The best QSPR models presented in this study showed good internal (range of : 89–92%) as well as external predictivity (range of Q ext²: 75–89%). The AD of the models was analysed by the leverage approach, and was represented graphically in the Williams graph. †Presented at the 13th International Workshop on QSARs in the Environmental Sciences (QSAR 2008), 8–12 June 2008, Syracuse, USA.

Linear QSAR regression models for the prediction of bioconcentration factors by physicochemical properties and structural theoretical molecular descriptors

The development of QSAR models useful for the prediction of fish bioconcentration factor (BCF) for a wide range of different chemical classes is crucial for the assessment and prioritisation of potentially persistent bioaccumulative and toxic substances. In this study we present QSAR models for BCF developed on a wide range of chemical structural classes of environmental and toxicological interest (such as dyes and various chlorinated and brominated compounds). The aim is to provide valid QSAR models, statistically validated for predictivity, for the prediction of BCF in general, but also for problematical chemical classes such as highly hydrophobic chemicals. Several descriptors, calculated by different commercially available software packages, have been employed in order to take into account relevant information provided by physicochemical properties (octanol/water partition coefficient and water solubility) and molecular features (structural and quantum-chemical molecular descriptors). The best descriptor subsets for the models were selected using the Genetic Algorithm-Variable Subset Selection strategy (GA-VSS) and calculations were performed by ordinary least squares regression. Starting from a data set of 640 compounds (log K ow range from −2.34 to 12.66), we developed linear QSARs, firstly for a data set of 620 compounds (log K ow range from −2.34 to 10.35) and secondly specifically for 87 highly hydrophobic chemicals (log K ow range from 6.00 to 10.35). All these models have been statistically validated (both internally by cross-validation and bootstrap and externally, by “a priori” splitting of available data by Kohonen Map-ANN in training and prediction sets) and their structural chemical domain has been verified by the leverage approach.

Statistical external validation and consensus modeling: A QSPR case study for Koc prediction

The soil sorption partition coefficient (log K oc) of a heterogeneous set of 643 organic non-ionic compounds, with a range of more than 6 log units, is predicted by a statistically validated QSAR modeling approach. The applied multiple linear regression (ordinary least squares, OLS) is based on a variety of theoretical molecular descriptors selected by the genetic algorithms-variable subset selection (GA-VSS) procedure. The models were validated for predictivity by different internal and external validation approaches. For external validation we applied self organizing maps (SOM) to split the original data set: the best four-dimensional model, developed on a reduced training set of 93 chemicals, has a predictivity of 78% when applied on 550 validation chemicals (prediction set). The selected molecular descriptors, which could be interpreted through their mechanistic meaning, were compared with the more common physico-chemical descriptors log K ow and log S w. The chemical applicability domain of each model was verified by the leverage approach in order to propose only reliable data. The best predicted data were obtained by consensus modeling from 10 different models in the genetic algorithm model population.

Validation of a QSAR model for acute toxicity†

In the present study, a quantitative structure – activity relationship (QSAR) model has been developed for predicting acute toxicity to the fathead minnow (Pimephales promelas), the aim being to demonstrate how statistical validation and domain definition are both required to establish model validity and to provide reliable predictions. A dataset of 408 heterogeneous chemicals was modelled by a diverse set of theoretical molecular descriptors by using multivariate linear regression (MLR) and Genetic Algorithm – Variable Subset Selection (GA-VSS). This QSAR model was developed to generate reliable predictions of toxicity for organic chemicals not yet tested, so particular emphasis was given to statistical validity and applicability domain. External validation was performed by using OECD Screening Information Data Set (SIDS) data for 177 High Production Volume (HPV) chemicals, and a good predictivity was obtained ( = 72.1). The model was evaluated according to the OECD principles for QSAR validation, and compliance with all five principles was established. The model could therefore be useful for the regulatory assessment of chemicals. For example, it could be used to fill data gaps within its chemical domain and contribute to the prioritization of chemicals for aquatic toxicity testing. †Presented at CMTPI 2005: Computational Methods in Toxicology and Pharmacology Integrating Internet resources (Shanghai, China, October 29–November 1 2005).

Prediction of ozone tropospheric degradation rate constant of organic compounds by using artificial neural networks

Ozone tropospheric degradation of organic compound is very important in environmental chemistry. The lifetime of organic chemicals in the atmosphere can be calculated from the knowledge of the rate constant of their reaction with free radicals such as OH and NO 3 or O 3. In the present work, the rate constant for the tropospheric degradation of 137 organic compounds by reaction with ozone, the least widely and successfully modeled degradation process, are predicted by quantitative structure activity relationships modeling based on a variety of theoretical descriptors, which screened and selected by genetic algorithm variable subset selection procedure. These descriptors which can be used as inputs for generated artificial neural networks are; HOMO–LUMO gap, number of double bonds, number of single bonds, maximum net charge on C atom, minimum (>0.1) bond order of C atom and Minimum e–e repulsion of H atom. After generation, optimization and training of artificial neural network, network was used for the prediction of log KO 3 for the validation set. The root mean square error for the neural network calculated log KO 3 for training, prediction and validation set are 0.357, 0.460 and 0.481, respectively, which are smaller than those obtained by multiple linear regressions model (1.217, 0.870 and 0.968, respectively). Results obtained reveal the reliability and good predictivity of neural network model for the prediction of ozone tropospheric degradations rate constant of organic compounds.

Statistically Validated QSARs, Based on Theoretical Descriptors, for Modeling Aquatic Toxicity of Organic Chemicals in Pimephales promelas (Fathead Minnow)

The use of Quantitative Structure-Activity Relationships in assessing the potential negative effects of chemicals plays an important role in ecotoxicology. (LC50)(96h) in Pimephales promelas (Duluth database) is widely modeled as an aquatic toxicity end-point. The object of this study was to compare different molecular descriptors in the development of new statistically validated QSAR models to predict the aquatic toxicity of chemicals classified according to their MOA and in a unique general model. The applied multiple linear regression approach (ordinary least squares) is based on theoretical molecular descriptor variety (1D, 2D, and 3D, from DRAGON package, and some calculated logP). The best combination of modeling descriptors was selected by the Genetic Algorithm-Variable Subset Selection procedure. The robustness and the predictive performance of the proposed models was verified using both internal (cross-validation by LOO, bootstrap, Y-scrambling) and external statistical validations (by splitting the original data set into training and validation sets by Kohonen-artificial neural networks (K-ANN)). The model applicability domain (AD) was checked by the leverage approach to verify prediction reliability.

Total ranking models by the genetic algorithm variable subset selection (GA-VSS) approach for environmental priority settings.

Total order ranking (TOR) strategies, which are mathematically based on elementary methods of discrete mathematics, seem to be attractive and simple tools for performing data analysis. Moreover order-ranking strategies seem to be a very useful tool not only to perform data exploration but also to develop order ranking models, a possible alternative to conventional quantitative structure-activity relationship (QSAR) methods. In fact, when data material is characterised by uncertainties, order methods can be used as alternative to statistical methods such as multilinear regression (MLR), because they do not require specific functional relationships between the independent and dependent variables (responses). A ranking model is a relationship between a set of dependent attributes, experimentally investigated, and a set of independent attributes, i.e. model attributes, which are calculated attributes. As in regression and classification models, the variable selection model is one of the main steps in finding predictive models. In this work the genetic algorithm-variable subset selection (GA-VSS) approach is proposed as the variable selection method for searching for the best ranking models within a wide set of variables. The models based on the selected subsets of variables are compared with the experimental ranking and evaluated by the Spearman's rank index. A case study application is presented on a TOR model developed for polychlorinated biphenyl (PCB) compounds, which have been analysed according to some of their physicochemical properties which play an important role in their environmental impact.

Validated QSAR prediction of OH tropospheric degradation of VOCs: splitting into training-test sets and consensus modeling.

The rate constant for hydroxyl radical tropospheric degradation of 460 heterogeneous organic compounds is predicted by QSAR modeling. The applied Multiple Linear Regression is based on a variety of theoretical molecular descriptors, selected by the Genetic Algorithms-Variable Subset Selection (GA-VSS) procedure. The models were validated for predictivity by both internal and external validation. For the external validation two splitting approaches, D-optimal Experimental Design and Kohonen Artificial Neural Networks (K-ANN), were applied to the original data set to compare the two methodologies. We emphasize that external validation is the only way to establish a reliable QSAR model for predictive purposes. Predicted data by consensus modeling from different models are also proposed.

Prediction of aromatic amines mutagenicity from theoretical molecular descriptors

In the present research the mutagenicity data (Ames tests TA98 and TA100) for various aromatic and heteroaromatic amines, a data set extensively studied by other quantitative structure–activity relationship (QSAR)-authors, have been modeled by a wide set of theoretical molecular descriptors using linear multivariate regression (MLR) and genetic algorithm–variable subset selection (GA–VSS). The models have been calculated on a subset of compounds selected by a D-optimal experimental design. Moreover, they have been validated by both internal and external validation procedures showing satisfactory predictive performance. The models proposed here can be useful in predicting data and setting a testing priority for those compounds for which experimental data are not available or are not yet synthesized.

QSAR Prediction of Ozone Tropospheric Degradation

AbstractThe lifetime of organic chemicals in the atmosphere can be calculated from a knowledge of the rate constant of their reaction with free radicals (OH., NO3.) and O3. The rate constant for the tropospheric degradation of 125 organic compounds by reaction with ozone, the least widely and successfully modelled degradation process, is predicted here by MLR‐QSAR modelling based on a variety of theoretical molecular descriptors, selected by the Genetic Algorithm‐Variable Subset Selection (GA‐VSS) procedure. The proposed models, checked for their reliability and robustness, have good predictivity, verified by internal (Q2LMO(50%)=82–88%) and also external validation (Q2EXT =90%). The best splitting of the original data set into representative training and test sets has been obtained by the Experimental Design approach. The model applicability domain was always verified by the leverage approach. The average root‐mean square error (RMS) for the prediction of log kO3 was 0.73, similar to the typical range of experimental error.

QSAR study on the tropospheric degradation of organic compounds

In the environmental risk assessment of organic chemicals, persistence is of particular importance as it may lead to adverse effects. The reaction of chemicals with OH and N0 3 radicals and ozone are the main abiotic degradation processes in the troposphere, so an upper limit of the atmospheric persistence of chemicals is assessed by determining their reaction rate constants with OH . and NO 3 . and O 3. Statistical models predicting the oxidation rate constants with OH . and NO 3 . for many heterogeneous compounds have been developed by the QSAR/QSPR ( Quantitative Structure-Activity/Property Relationships) approach; the structural representation of the compounds was realised using different kinds of molecular descriptors ( structural, topological, empirical and WHIM descriptors). In addition, Kohonen neural networks (K-ANN) and the GA-VSS (Genetic Algorithm Variable Subset Selection) strategy were respectively used to select the most representative training set and the best descriptor subset. The predictive capability of the models on k OH and k NO3 has been checked and appears to be satisfactory. Finally, the oxidation rate constants for some chemicals of concern were analysed in the Principal Component space in order to rank these chemicals according to their tropospheric degradability.