Support vector machines in the prediction of mutagenicity of chemical compounds
Thomas Ferrari, Giuseppina C. Gini, Emilio Benfenati · 2009
In this paper we introduce the problem of predicting the mutagenic toxicity property of chemical compounds and we discuss how this can be partially formulated as a computational intelligence problem. Then we develop a statistical model based on a selected set of descriptors of the molecular structure. The classifier, that we derived from SVM methods, outperforms the available methods in performance and simplicity.