Bihai Zhao, Jianxin Wang* and Fang-Xiang Wu Pages 1120 - 1131 ( 12 )
Predicting functions of proteins is a key issue in the post-genomic era. Some experimental methods have been designed to predict protein functions. However, these methods cannot accommodate the vast amount of sequence data due to their inherent difficulty and expense. To address these problems, a lot of computational methods have been proposed to predict the function of proteins. In this paper, we provide a comprehensive survey of the current techniques for computational prediction of protein functions. We begin with introducing the formal description of protein function prediction and evaluation of prediction methods. We then focus on the various approaches available in categories of supervised and unsupervised methods for predicting protein functions. Finally, we discuss challenges and future works in this field.
Protein-protein interaction, protein function prediction, neural network, frequent pattern, support vector machine, heterogeneous data fusion, functional similarity.
Department of Mathematics and Computer Science, Changsha University, Changsha, 410022, Computer Building 303, School of Information Science and Engineering, Central South University, Changsha 410083, School of Information Science and Engineering, Central South University, Changsha, 410083