The accurate QSPR models to predict the bioconcentration factors of nonionic organic compounds based on the heuristic method and support vector machine Huanxiang Liu a , Xiaojun Yao a, * , Ruisheng Zhang a,b , Mancang Liu a , Zhide Hu a , Botao Fan c a Department of Chemistry, Lanzhou University, Lanzhou 730000, China b Department of Computer Science, Lanzhou University, Lanzhou 730000, China c Universite ´ Paris 7-Denis Diderot, ITODYS 1, Rue Guy de la Brosse, 75005 Paris, France Received 15 April 2005; received in revised form 8 July 2005; accepted 12 August 2005 Available online 14 October 2005 Abstract The heuristic method (HM) and support vector machine (SVM) were used to build the linear and nonlinear quant- itive structure–property relationship (QSPR) models for the prediction of the fish bioconcentration factors (BCF) for 122 diverse nonionic organic chemicals using the three descriptors calculated from the molecular structure alone and selected by HM. Both the linear and nonlinear model can give very satisfactory prediction results: the square of cor- relation coefficient R 2 was 0.929 and 0.953, the root mean square (RMS) error was 0.404 and 0.331, respectively for the whole dataset. The prediction result of the SVM model is better than that obtained by heuristic method, which proved SVM was a useful tool in the prediction of the BCF. At the same time, the HM model showed the influencing degree of different molecular descriptors on bioconcentration factors and then could improve the understanding for the bioconcentration mechanism of organic pollutants from molecular level. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Bioconcentration factors; QSPR; Support vector machine; Heuristic method 1. Introduction As a large amount of organic contaminants spread widely in air, water and soil, an environmental problem is increasingly attracting public awareness all over the world. Particularly, plenty of halogenated organic com- pounds, polychlorinated nitro benzenes and phenols, and alkyl benzenes and phenols have been discharged into the environment. Most of them are persistent in the environment and also resistant to chemical degrada- tion, then show a tendency to accumulate in biota, soils, and sediments and are also dispersed in the atmosphere, giving them long half-lives in the environment (Wania and Mackay, 1996). Long-lived organic contaminants can biomagnify in food chains and may ultimately cause 0045-6535/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2005.08.031 * Corresponding author. Tel.: +86 931 8912578; fax: +86 931 8912582. E-mail address: liuhx1003@yahoo.com.cn (X. Yao). Chemosphere 63 (2006) 722–733 www.elsevier.com/locate/chemosphere