Predictions of chromatographic retention indices of alkylphenols with support vector machines and multiple linear regression

Abstract

In this study, quantitative structure-retention relationship (QSRR) was used for the prediction of Kováts retention indices of 180 alkylphenols and their derivatives using the multiple linear regression (MLR) and support vector machine (SVM). After the calculation of some molecular descriptors for all molecules, the data set was randomly divided into training and test sets. The diversity of training and test sets was examined by molecular diversity validation test. Then stepwise MLR was used for the selection of the most important descriptors and development of MLR models. Descriptors which appeared in these QSRR models are number of H atoms, relative number of O atoms, Balaban index, relation yz-shadow/yz-rectangle and partial charges hydrogen bond donor atoms HDCA 2 index. These descriptors were used as inputs for developing the SVM model. After optimizing the SVM parameters, it was used for the calculation of chromatographic retention of interest molecules. The values of SE in calculation of Kováts retention indices for training and test sets are 0.34 and 0.63, respectively, for MLR model and 0.35 and 0.63, respectively, for SVM model. The overall values of average absolute relative error were 13.24 and 13.83 for MLR and SVM models, respectively. In addition, the cross-validation tests were performed to further examine the obtained model. IntroductionAlkylphenols (APs) are a family of chemicals that are mainly used as raw materials in the production of varieties of industrial products such as surfactants, detergents, phenolic resins, polymer additives and lubricants [1,2]. Long-chained branched para-substituted isomers of APs are known to have oestrogenic activity and therefore represent an environmental problem [3][4][5]. APs are present in raw oil and those of shorter alkyl chain may enter into the environment via produced water. Produced water is defined as the water that comes up with oil and gas from sea bed reservoirs, separated on the platform from the oil and discharged into the sea, from offshore oil installations [6]. Due to biohazards in aquatic environment of APs, the detection and quantification of them are very important. Few methods have been developed to determine APs in the aquatic environment and most of them invariably comprise three steps, namely, extraction/clean-up/extract evaporation, chromatographic separation and mass spectroscopic detection [7][8][9][10][11][12][13]. The identification of these compounds by chromatography was done on the basis of their retention parameters. On a GC retention index scale, the retention of a compound is described in relation to the retention of a series of homologues. The most common retention system is the Kováts retention indices [14,15], where n-alkanes are applied as reference compounds. Retention indices are independent of carrier gas flow, column dimensions and phase ratios [16]. Special retention index for alkylphenol Abbreviations: AARD, average absolute relative deviation; AP, alkylphenol; APRI X , retention index for alkylphenol; BI, Balaban...

Prediction of gas chromatographic retention indices of some amino acids and carboxylic acids from their structural descriptors

“…Liao et al developed a method of molecular structural characterization called hydrogen-association classified molecular electronegativity-distance vector (H-MEDV) to describe the retention of 106 oxygen-containing organic compounds [19]. Also, some QSRR models were developed to study the chromatographic retention of some organic nucleophiles [20], alkyl phenols [21] and some organic pollutants [22] in our laboratory. In the present work, we try to establish a new QSRR model for predicting retention indices of AAs and CAs in inherited metabolic disorders by using MLR, artificial neural network (ANN) and support vector machine (SVM) techniques as feature mapping techniques.…”

Multiple Linear Regression: An Overview with Analytical and Physico-Chemical Applications

“…The field of prediction is also an important area, e.g. chromatographic retention indices [96,99,100,120,135], acute toxicity of phenol derivatives [74], some physico-chemical properties [84], apparent volume of distribution [108], partition coefficients [122], acidity constant values [123] fuel ignition quality by NMR [76], models for ginsenosides [100], and performance of MLR [139]. Solution kinetics [73,142], modelling and optimization on ionisable compounds [111] and acid-base behaviour of solvent effects [117], analysis of IR and NIR-FT spectra [77,85,90], UV absorption spectra [81], correction of spectral interferences (matrix effect) [87,93,129] on molecular absorption and atomic emission techniques, and multicomponent spectrophotometric analysis of mixtures [94,95,107,146] have also been the subject of study.…”

Quantitative structure-retention relationship prediction of Kováts retention index of some organic acids

“…Moreover, Kováts retention indices of 180 alkylphenols and their derivatives were predicted by using the MLR and support vector machine (SVM) in our laboratory [16]. Other works about QSRR prediction of chromatographic parameters in our laboratory can be found in the references [17][18][19][20][21]. In continuation of our previous investigations, in the present work, we try to establish a new QSRR model for predicting retention indices of AAs and CAs in IMDs by using MLR, artificial neural network (ANN) and SVM techniques as feature mapping techniques.…”