Detection of functional modules from protein interaction networks
Detection of functional modules from protein interaction networks
复制标题
DOI：
10.1002/prot.10505
复制
发表时间：
2004-01-01
影响因子：
2.9
通讯作者：
Ouzounis, CA
中科院分区：
文献类型：
--
作者：
Pereira-Leal, JB;Enright, AJ;Ouzounis, CA
复杂的细胞过程是模块化的，并通过功能模块的协同作用来完成（Ravasz 等人，Science 2002；297：1551-1555；Hartwell 等人，Nature 1999；402：C47-52）。这些模块包含参与常见基本生物功能的基因或蛋白质组。功能基因组学的一个重要且很大程度上尚未解决的目标是从全基因组信息（例如转录谱或蛋白质相互作用）中识别功能模块。为了应对蛋白质相互作用数据不断增加的数量和复杂性（Bader 等人，Nucleic Acids Res 2001；29:242-245；Xenarios 等人，Nucleic Acids Res 2002；30:303-305），需要在这些密集连接的相互作用网络中进行模式发现的新自动化方法（Ravasz 等人，Science） 2002；297：1551-1555；Bader 和 Hogue，Nat Biotechnol 2002；20：991-997；Snel 等人，Proc Natl Acad Sci USA 2002；99：5890-5895）。在这项研究中，我们通过使用完全自动化和无监督的图聚类算法，成功地从已知的酿酒酵母蛋白质相互作用网络中分离出 1046 个功能模块，涉及 8046 个单独的成对相互作用。这种系统生物学方法能够检测许多众所周知的蛋白质复合物或生物过程，而无需参考任何其他信息。我们使用广泛的统计验证程序来确定检测到的模块的生物学意义，并探索这种复杂的、模块化相互作用的分层网络，从中可以推断出路径。 (C) 2003 Wiley-Liss, Inc.
Complex cellular processes are modular and are accomplished by the concerted action of functional modules (Ravasz et al., Science 2002;297:1551-1555; Hartwell et al., Nature 1999;402: C47-52). These modules encompass groups of genes or proteins involved in common elementary biological functions. One important and largely unsolved goal of functional genomics is the identification of functional modules from genomewide information, such as transcription profiles or protein interactions. To cope with the ever-increasing volume and complexity of protein interaction data (Bader et al., Nucleic Acids Res 2001;29:242-245; Xenarios et al., Nucleic Acids Res 2002;30:303-305), new automated approaches for pattern discovery in these densely connected interaction networks are required (Ravasz et al., Science 2002;297:1551-1555; Bader and Hogue, Nat Biotechnol 2002;20:991-997; Snel et al., Proc Natl Acad Sci USA 2002;99:5890-5895). In this study, we successfully isolate 1046 functional modules from the known protein interaction network of Saccharomyces cerevisiae involving 8046 individual pair-wise interactions by using an entirely automated and unsupervised graph clustering algorithm. This systems biology approach is able to detect many well-known protein complexes or biological processes, without reference to any additional information. We use an extensive statistical validation procedure to establish the biological significance of the detected modules and explore this complex, hierarchical network of modular interactions from which pathways can be inferred. (C) 2003 Wiley-Liss, Inc.