Jeeva: Enterprise Grid Enabled Web Portal for Protein Secondary Structure Prediction

Abstract

Abstract-This paper presents a Grid portal for protein secondary structure prediction developed by using services of Aneka, a .NET-based enterprise Grid technology. The portal is used by research scientists to discover new prediction structures in a parallel manner. An SVM (Support Vector Machine)-based prediction algorithm is used with 64 sample protein sequences as a case study to demonstrate the potential of enterprise Grids. I. INTRODUCTIONThe structure of protein plays a key role in the structurebased design of drugs for the treatment of various diseases. However, it is still a challenge to find out protein structure based on its sequence, and the dependence on experimental methods may not yield protein structures fast enough to keep up with the requirement of current industry. Fortunately, the energy landscape theory [24] enables a framework for the development of algorithms to predict the structure of unknown proteins based on their sequence, which is known as protein structure prediction.From the perspective of computer science, protein structure prediction is a computing intensive task [5]. Since the prediction of protein structure is a complex task, it is usually sub-divided into two phases. The first one is secondary structure prediction and the second one is super secondary structure prediction, leading to tertiary structure, i.e., the specific atomic positions in three-dimensional space. As the first phase of protein structure prediction, accurate secondary structure prediction is a key element for correctly acquiring tertiary structure.A large number of algorithms [2][6][9][11] have been proposed for protein secondary structure prediction. To facilitate the collaboration between protein scientists across the world, it is a necessity for researchers to share their algorithms and results with colleagues dispersed at different geographical locations. Furthermore, to speed up the process of finding out new protein structures, we need a proper computational platform which simplifies the development of new prediction algorithms and improves the efficiency at the