Disease gene identification using reliable robust classifier

Abstract

Identification of genes causing the diseases is a major challenging problem towards diagnosing and providing treatment in a earlier manner. Many motivating methodologies are being proposed for the identification of disease genes. Generally, the unique variation among the previously proposed methodologies depend on the prior knowledge, also machine learning methodologies utilized for identifying. Identification of disease gene is normally observed as two class classification issue. Nature of information generates a key issue which can have an effect on results. In this research work, reliable robust classifier (RRC) based on dual simplex concept has been proposed to allocate a genes to a single disease class. RRC classifies the genes of classes into vertices of dimension dual simplex which results in -class classification turn out to be class task. Since there exist no benchmark method to characterize the genes that have-diseases and not-have-diseases, this research work utilizes support vector machine to predict it. The results of experiments clearly demonstrate the effectiveness of the method with better precision, recall, and F-measure respectively.