Jiang Rui, Tu Zhidong, Chen Ting, Sun Fengzhu
Molecular and Computational Biology Program, University of Southern California, Los Angeles, CA 90089, USA.
Proc Natl Acad Sci U S A. 2006 Jun 20;103(25):9404-9. doi: 10.1073/pnas.0507841103. Epub 2006 Jun 12.
Network motifs have been identified in a wide range of networks across many scientific disciplines and are suggested to be the basic building blocks of most complex networks. Nonetheless, many networks come with intrinsic and/or experimental uncertainties and should be treated as stochastic networks. The building blocks in these networks thus may also have stochastic properties. In this article, we study stochastic network motifs derived from families of mutually similar but not necessarily identical patterns of interconnections. We establish a finite mixture model for stochastic networks and develop an expectation-maximization algorithm for identifying stochastic network motifs. We apply this approach to the transcriptional regulatory networks of Escherichia coli and Saccharomyces cerevisiae, as well as the protein-protein interaction networks of seven species, and identify several stochastic network motifs that are consistent with current biological knowledge.
网络基序已在众多科学学科的广泛网络中被识别出来，并被认为是大多数复杂网络的基本构建单元。然而，许多网络存在内在和/或实验上的不确定性，应被视为随机网络。因此，这些网络中的构建单元也可能具有随机性质。在本文中，我们研究了源自相互相似但不一定相同的互连模式家族的随机网络基序。我们为随机网络建立了一个有限混合模型，并开发了一种期望最大化算法来识别随机网络基序。我们将这种方法应用于大肠杆菌和酿酒酵母的转录调控网络，以及七个物种的蛋白质-蛋白质相互作用网络，并识别出了几个与当前生物学知识一致的随机网络基序。
