Ming Fang, Xiujuan Lei* and Ling Guo
Essential proteins play important roles in the survival or reproduction of an organism and support the stability of the system. Essential proteins compose the minimum set of proteins absolutely required to maintain a living cell. The identification of essential proteins is a very important topic not only for a better comprehension of the minimal requirements for cellular life, but also for a more efficient discovery of the human disease genes and drug targets. Traditionally, the experimental identification of essential proteins is complex, usually requires great time and expense. With the cumulation of high-throughput experimental data, many computational methods that make useful complements to experimental methods have been proposed to identify essential proteins. In addition, the ability to rapidly and precisely identify essential proteins is of great significance for discovering disease genes and drug design, and has great potential for applications in basic and synthetic biology research. The aim of this paper is to provide a review on the identification of essential proteins and genes focusing on the current developments of different types of computational methods, point out some progress and limitations of existing methods, and the challenges and directions for further research are discussed.
Essential proteins, essential genes, machine learning algorithms, computational techniques, ensemble methods, protein-protein interaction network
School of Computer Science, Shaanxi Normal University, Xi`an 710119, School of Computer Science, Shaanxi Normal University, Xi`an 710119, College of Life Sciences, Shaanxi Normal University, Xi`an 710119