Prediction of the Odor Thresholds of Oxygen and Nitrogen Containing Heterocycles Using the Quantitative Structure-Property Relationship Approach

Xuan Xu, Feng Luan*, Huitao Liu and Yuan Gao

Department of Applied Chemistry, Yantai University, Yantai 264005, P.R. China

*Corresponding author: Fax: +86 535 6902063; Tel: +86 535 6902063; E-mail: fluan@sina.com

Abstract

Quantitative structure-property relationship (QSPR) models are developed to correlate the odor thresholds of 50 oxygen and nitrogen containing heterocycles from their molecular structures. Three statistic methods including multiple linear regression (MLR), non-linear radial basis function neural network (RBFNN) and support vector machine (SVM) are performed to build the models. A six-descriptor equation with the squared correlation coefficient (R2) of 0.8012 and root mean square error (RMS) of 1.0011 were obtained for the training set and R2 = 0.648, RMS = 1.7165 for the external test set. The radial basis function neural network model gave better results: R2 = 0.8767, RMS = 0.7165 for the training set and R2 = 0.7746, RMS = 1.3570 for the external test set. The SVM model gave similar results to multiple linear regression, that is, R2 = 0.8023, RMS = 0.9271 for the training set and R2 = 0.7033 and RMS = 1.5888 for the test set. The aim of the paper is to provide an easy, direct and relatively accurate way to estimate the odor thresholds.

Keywords

Odor thresholds, Quantitative structure-property relationship, Multiple linear regression, Radial basis function neural network, Support vector machines.

   View Article PDF File Under a Creative Commons License
 
  • Asian J. Chem. /
  •  2012 /
  •  24(9) /
  •  pp 3842-3848