Mutagenicity prediction using chemoinformatics methods

Abstract

The objective of this study is to construct a model which can predict mutagenicity of organic compounds with high accuracy. For this end, we propose a novel ensemble modeling method in which a lot of support vector machine (SVM) models are constructed as a sub-model and integrated to predict mutagenicity. For constructing sub-models, a part of data matrix is randomly selected from the original data matrix. After the construction of sub-models, a certain number of models which have high accuracy rate are selected and integrated to predict mutagenicity. We constructed an ensemble model using a data set of reverse mutation test which was collected by Hansen et al. to estimate the proposed method. As a result, the ensemble models with accuracy rate of 79.6% and 80.9% were successfully obtained. Moreover, we carried out reverse mutation test in order to verify correctness of the database. Five compounds, which are registered as negative in database but are predicted as positive by our model, were systematically selected and mutagenicity of these compounds was estimated. As a result, it is found that three of five compounds have mutagenicity.