Quantitative Structure Activity Relationship Models For The Prediction Of Rat Lethal Dose 50% Of Aldehydes Based On Heuristic Method And Support Vector Machine
Zhang Ke-Jun,Sun Shou-Qian,Chai Chun-Lei,Si Hong-Zong
DOI: https://doi.org/10.3321/j.issn:0253-3820.2007.09.005
IF: 1.193
2007-01-01
Chinese Journal of Analytical Chemistry
Abstract:The quantitative structure activity relationship (QSPR) models of rat lethal dose 50% (LD50) of aldehydes were constructed by heuristic method (HM) and support vector machine (SVM). The structures of rat LD50 of 88 diverse aldehydes were drawn in the software chemOffice 2004. Semi-empirical method was used to optimize the structures. Basis on this step, five classes of descriptors of constitutional, topological, geometrical, electrostatic and quantum chemical parameters were obtained by CODDESA software. 8 Descriptors of RPCG Relative positive charge, number of 0 atoms, min nueleoph. react. index for a C atom, relative number of double bonds, number of C atoms, max exchange energy for a C-H bond, max SIGMA-SIGMA bond order and number of double bonds were selected by HM. The HM and SVM were utilized to construct the linear and nonlinear prediction models, leading to a good cross-validation correlation coefficient (R-CV) of 0. 90 and 0. 93, respectively. The specific information described by the heuristic linear model could give some insights into the factors that are likely to govern the rat LD50 of the compounds and the prediction results of the nonlinear SVM model seem to be better than that of HM.