摘要
本文研究了组织特异性蛋白质复合体的识别问题.利用蛋白质相互作用网络数据以及组织特异性基因表达数据构建组织特异性蛋白网络,利用多种代表性聚类算法对该网络进行聚类,并利用非负矩阵分解对聚类结果进行合并聚类,得到了组织特异性蛋白质复合体.结果表明,聚类效果得到明显提升,并且能识别出组织特异性蛋白质复合体.
In this paper, we study the identification problem of tissue-specific protein complexes. By using a variety of typical clustering algorithm to cluster the network, we construct a tissue-specific protein-protein interaction network based on the protein-protein interaction net- works as well as the tissue-specific gene expression data, then merge the results with non-negative matrix factorization model to obtain tissue-specific protein complexes. The results show that clustering effect has been significantly improved, and can identify tissue-specific protein complexes.
出处
《数学杂志》
北大核心
2017年第5期1093-1100,共8页
Journal of Mathematics
基金
国家自然科学基金资助(11275259)
国家自然科学基金资助(91330113)
关键词
蛋白质相互作用网络
复合体识别
组织特异性
非负矩阵分解
protein-protein interaction networks
complexes identification
tissue-specific
non-negative matrix factorization
