生物医学工程研究2009，Vol.28Issue(2)：87-90,4.
基于K近邻的蛋白质功能的预测方法
Prediction of Protein Functions based on K Nearest Neighbors Method
摘要
Abstract
In the post-genomic era, protein function annotation is one of the most important problems. In this paper, a new method, based on K Nearest Neighbors method, was proposed to predict protein functions from protein-protein interactions. The similarity of protein annotation environments between the unannotated protein and all the annotated proteins were computed and the K proteins, which were most similar to the unannotated protein, were selected, then a weighted average of the functions of the K proteins was used as the functions estimate for the unannotated protein. The results, which is obtained by testing on the two large scale data sets constructed for Saccharomyces, show that the proposed method can predict protein functions effectively and outperforms some existing approaches.关键词
蛋白质相互作用/蛋白质功能预测/K近邻算法/相似性/加权方法Key words
Protein-protein interactions/Protein function prediction/K nearest neighbor method/Similarity/Weighted method分类
生物科学引用本文复制引用
倪青山,王正志,黎刚果,孟祥林..基于K近邻的蛋白质功能的预测方法[J].生物医学工程研究,2009,28(2):87-90,4.基金项目
国家自然科学基金项目(60471003) （60471003）
