Evaluation of data mining classifiers for prediction and classification of Glaucoma associated proteins

Abstract

The human’s gene expressions are the relative factors that determine the ability of the protein to cause disease. Predicting the ability of proteins to cause glaucoma by analyzing its amino acid composition and arrangements might play a crucial role in fighting many eye- related diseases. Though the availability of many specialized databases is committed to providing the protein sequence information, the limitation in practical exercises exists in determining the proteins that are associated with glaucoma. In this scenario, the development of computational methods such as machine learning approaches serve as time consuming alternatives. Thus in this paper, the classifier methods were evaluated to understand their ability in terms of sensitivity, specificity and accuracy in predicting the association of protein as glaucoma associated or not. This study revealed that SMO classifier (Support vector machine) excels in predicting the protein’s association with Glaucoma disease, based on its Amino Acid Composition (AAC) at five-fold cross-validation.