Machine learning-based quantitative structure–retention relationship models for predicting the retention indices of volatile organic pollutants

The chemical composition of the volatile fraction obtained by head-space solid phase microextraction (HS-SPME), single drop microextraction (SDME) and the essential oil obtained by cold-press from the peels of C. sinensis cv. valencia were analyzed employing gas chromatography-flame ionization detector (GC-FID) and gas chromatography-mass spectrometry (GC-MS). The main components were limonene (61.34 %, 68.27 %, 90.50 %), myrcene (17.55 %, 12.35 %, 2.50 %), sabinene (6.50 %, 7.62 %, 0.5 %) and ?-pinene (0 %, 6.65 %, 1.4 %) respectively obtained by HS-SPME, SDME and cold-press. Then a quantitative structure-retention relationship (QSRR) study for the prediction of retention indices (RI) of the compounds was developed by application of structural descriptors and the multiple linear regression (MLR) method. Principal components analysis was used to select the training set. A simple model with low standard errors and high correlation coefficients was obtained. The results illustrated that linear techniques such as MLR combined with a successful variable selection procedure are capable of generating an efficient QSRR model for prediction of the retention indices of different compounds. This model, with high statistical significance (R2 train = 0.983, R2 test = 0.970, Q2 LOO = 0.962, Q2 LGO = 0.936, REP(%) = 3.00), could be used adequately for the prediction and description of the retention indices of the volatile compounds.

Download Full-text

A quantitative structure-retention relationship for the prediction of retention indices of the essential oils of Ammoides atlantica

Journal of the Serbian Chemical Society ◽

10.2298/jsc100219076a ◽

2011 ◽

Vol 76 (6) ◽

pp. 891-902 ◽

Cited By ~ 7

Author(s):

Aberomand Azar ◽

Mehdi Nekoei ◽

Siavash Riahi ◽

Mohamad Ganjali ◽

Karim Zare

Keyword(s):

Genetic Algorithm ◽

Essential Oil ◽

Essential Oils ◽

Retention Indices ◽

Quantitative Structure ◽

Data Set ◽

Prediction Of Retention ◽

Structure Retention ◽

Oil Components ◽

Quantitative Structure Retention Relationship

A simple, descriptive and interpretable model, based on a quantitative structure-retention relationship (QSRR), was developed using the genetic algorithm-multiple linear regression (GA-MLR) approach for the prediction of the retention indices (RI) of essential oil components. By molecular modeling, three significant descriptors related to the RI values of the essential oils were identified. A data set was selected consisting of the retention indices for 32 essential oil molecules with a range of more than 931 compounds. Then, a suitable set of the molecular descriptors was calculated and the important descriptors were selected with the aid of the genetic algorithm and multiple regression method. A model with a low prediction error and a good correlation coefficient was obtained. This model was used for the prediction of the RI values of some essential oil components which were not used in the modeling procedure.

Download Full-text

Quantitative structure-retention relationship model for predicting retention indices of constituents of essential oils of Thymus vulgaris (Lamiaceae)

Journal of the Serbian Chemical Society ◽

10.2298/jsc180817010d ◽

2019 ◽

Vol 84 (4) ◽

pp. 405-416 ◽

Cited By ~ 3

Author(s):

Youssouf Driouche ◽

Djelloul Messadi

Keyword(s):

Essential Oils ◽

Statistical Significance ◽

Selection Procedure ◽

Randomization Test ◽

Retention Indices ◽

Molecular Structures ◽

Thymus Vulgaris ◽

Quantitative Structure ◽

Structure Retention ◽

Quantitative Structure Retention Relationship

In this paper, a quantitative structure?retention relationship (QSRR) model was developed for predicting the retention indices (log RI) of 36 constituents of essential oils. First, the chemical structure of each compound was sketched using HyperChem software. Then, molecular descriptors covering different information of molecular structures were calculated by Dragon software. The results illustrated that linear techniques, such as multiple linear regression (MLR), combined with a successful variable selection procedure are capable of generating an efficient QSRR model for predicting the retention indices of different compounds. This model, with high statistical significance (R2 = 0.9781, Q2 LOO = 0.9691, Q2 ext = 0.9546, Q2 L(5)O = 0.9667, F = 245.27), could be used adequately for the prediction and description of the retention indices of other essential oil compounds. The reliability of the proposed model was further illustrated using various evaluation techniques: leave-5-out cross-validation, bootstrap, randomization test and validation through the test set.

Download Full-text