ISSN print edition: 0366-6352
ISSN electronic edition: 1336-9075
Registr. No.: MK SR 9/7

Published monthly

QSRR prediction of gas chromatography retention indices of essential oil components

Yovani Marrero-Ponce, Stephen J. Barigye, María E. Jorge-Rodríguez, and Trang Tran-Thi-Thu

Universidad San Francisco de Quito (USFQ), Grupo de Medicina Molecular y Traslacional (MeM&T), Quito, Ecuador



Abstract: A comprehensive and largest (to the best of our knowledge) database of 791 essential oil components (EOCs) with corresponding gas chromatographic retention properties has been built. With this data set, Quantitative structure–retention relationship (QSRR) models for the prediction of the Kováts retention indices (RIs) on the non-polar DB-5 stationary phase have been built using the DRAGON molecular descriptors and the regression methods: multiple linear regression (MLR) and artificial neural networks (ANN). The obtained models demonstrate good performance, evidenced by the satisfactory statistical parameters for the best MLR (R 2 = 96.75% and (Q_{ ext{ext}}^{2}) = 98.0%) and ANN (R 2 = 97.18% and (Q_{ ext{ext}}^{2}) = 98.4%) models, respectively. In addition, the built models provide information on the factors that influence the retention of EOCs over the DB-5 stationary phase. Comparisons of the statistical parameters for the QSRR models in the present study with those reported in the literature demonstrate comparable to superior performance for the former. The obtained models constitute valuable tools for the prediction of RIs for new EOCs whose experimental data are undetermined.

Keywords: Gas chromatography ; Retention index ; Essential oil ; Quantitative structure–retention relationships ; Multiple linear regression ; Artificial neural networks 

Full paper is available at

DOI: 10.1007/s11696-017-0257-x


Chemical Papers 72 (1) 57–69 (2018)

Thursday, May 23, 2024

SCImago Journal Rank 2021
European Symposium on Analytical Spectrometry ESAS 2022
© 2024 Chemical Papers