Towards A Multi Agent System Based Data Mining For Proteins Prediction And Classification

Abstract To understand the structure function paradigm in this paper a new algorithm for proteins classification and prediction is proposed. It uses multi agent system technique that represents a new paradigm for conceptualizing designing and implementing software systems to predict and classify the protein structures. For classifying the proteins support vector machine SVM has been developed to extract feature from the proteins sequences. This paper describes a method for predicting and classifying secondary structure of proteins. Support vector machine SVM modules were developed using multi-agent system principle for predicting the proteins and its function and achieved maximum accuracy specificity sensitivity of 92 94.09and 91.59 respectively. The proposed algorithm provide a good understanding for proteins structure which affect positively on biological science specially on understanding the behavior and the relationships between proteins.