Python(编程语言)
算法
计算机科学
数据挖掘
蛋白质检测
机器学习
纳米技术
操作系统
材料科学
作者
Milica Jaguzovic,Milana Grbić,Marko Đukanović,Dragan Matić
标识
DOI:10.1109/infoteh53737.2022.9751314
摘要
Community detection is of a major interest in network analysis. In this study several overlapping community detection algorithms are applied on different protein-protein interactions (PPIs) networks (BioGRID, String and WI-PHI) in order to examine their capability to identify protein complexes. Several community detection algorithms implemented in CDLIB Python library are examined. Obtained communities are further evaluated against four different gold standards of protein complexes from literature. The accuracy of the methods applied on the PPIs networks is examined by statistical measures designed to cope with overlapping partitions. The experimental results indicate that the community detection algorithms are more successful on BioGRID and WI-PHI networks, obtaining a relatively high accuracy in several cases.
科研通智能强力驱动
Strongly Powered by AbleSci AI