A Novel Approach For Classifying Protein Structures Based On Fuzzy Decision Tree

The processes in the organisms are influenced by protein molecules. Based on the functions of proteins, they can be used to design drugs for various diseases. There are methods that can be used to classify protein structures where the classification is manual done by domain experts, but they are not able to provide classification with reasonable speed, require intensive humans' effort and is time consuming. Thus, there is obvious necessity for rapid methods that would afford accurate classification of protein structures in an automated way. In this paper, we introduce an approach for classifying protein structures. First, for each protein its ray based descriptor is extracted, which gives evidence how the backbone of the protein is positioned with respect to the center of the protein. Then, a prediction model is created by using the fuzzy decision tree classifier. For evaluation, we used a part from the SCOP database, which holds information for the classification of proteins gathered in manual way. We present experimental results from the evaluation of the proposed approach.