Method for Identifying Essential Proteins by Key Features of Proteins in a Novel Protein-Domain Network

Frontiers in Genetics

Lei Wang¹  Yihong Tan¹  Zhiping Chen²  Linai Kuang³  Xin He³ 

[1] Applied Mathematics, Changsha University, Changsha, China;;College of Computer Engineering &College of Computer, Xiangtan University, Xiangtan, China;

关键词: essential proteins;    protein-protein network;    computational model;    domain-domain network;    protein-domain network;   

DOI  :  10.3389/fgene.2021.708162

来源: DOAJ

【 摘 要 】

In recent years, due to low accuracy and high costs of traditional biological experiments, more and more computational models have been proposed successively to infer potential essential proteins. In this paper, a novel prediction method called KFPM is proposed, in which, a novel protein-domain heterogeneous network is established first by combining known protein-protein interactions with known associations between proteins and domains. Next, based on key topological characteristics extracted from the newly constructed protein-domain network and functional characteristics extracted from multiple biological information of proteins, a new computational method is designed to effectively integrate multiple biological features to infer potential essential proteins based on an improved PageRank algorithm. Finally, in order to evaluate the performance of KFPM, we compared it with 13 state-of-the-art prediction methods, experimental results show that, among the top 1, 5, and 10% of candidate proteins predicted by KFPM, the prediction accuracy can achieve 96.08, 83.14, and 70.59%, respectively, which significantly outperform all these 13 competitive methods. It means that KFPM may be a meaningful tool for prediction of potential essential proteins in the future.

【 授权许可】

Unknown