Using Grey Model to Predict Protein Remote Homologous Family

2nd International Symposium on Resource Exploration and Environmental Science

生态环境科学

Xiao, Xuan^1 ; Cao, Weihua^1 ; Lin, Weizhong^1

Computer Department, Jing-De-Zhen Ceramic Institute, Jing-De-Zhen

333403, China^1

关键词: Benchmark datasets;    Computation methods;    Evolutionary relationships;    Grey relational analysis;    Protein structures;    Ranking methods;    Remote homology;    Remote homology detections;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/170/5/052019/pdf
DOI  :  10.1088/1755-1315/170/5/052019

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

Protein remote homology detection is the most basic and core problems of protein structure and function research. The purpose of protein remote homology detection to detect the remote evolutionary relationship between proteins by computation methods. At present, there are many methods for remote homology detection of proteins, but some methods are not very effective. In bioinformatics, it is urgent to further improve the performance of protein remote homology detection. In this study, we propose a new model called PHom-GRA to detect protein remote homology, which is the integration of various ranking methods via using grey relational analysis. Our experiment constructs benchmark dataset with lower homology, in which any two proteins have not more than 40% identity or homology. We achieve an ROC1 score of 0.7372 and an ROC50 score of 0.7968 in jack-knife test.