Prediction of protein essentiality by the improved particle swarm optimization

The essential protein is very important for understanding cellular critical activities and development. With the development of high throughput technology, how to identify the essential proteins from the protein interaction network has become a hot research topic in proteomics. A series of prediction methods have been proposed to infer the possibility of proteins to be essential by using the network topology. Therefore, it is necessary to develop an efficient method to detect the essential proteins considering both network topology and the biological attribute information. In this work, an effective method for essential proteins identification based on improved particle swarm optimization, named as EPPSO, is proposed. The method first constructs a weighted network by integrating the network topology characteristics and multi-source biological attribute information. To implement the PSO for essential protein identifying, we define the updating rules of the velocity vector and the positions of the particles. To estimate the essentiality of the nodes, we propose an index to measure the overall essentiality of the top-p essential proteins. The experimental results on yeast PPI data show that our algorithm is superior to other similar algorithms in terms of speed, accuracy and the number of essential proteins detected.