Grönlund Andreas
Department of Physics, Umeå University, 901 87 Umeå, Sweden.
Phys Rev E Stat Nonlin Soft Matter Phys. 2004 Dec;70(6 Pt 1):061908. doi: 10.1103/PhysRevE.70.061908. Epub 2004 Dec 20.
Two different types of directed networks are investigated, transcriptional regulation networks and neural networks. The directed network structure is studied and is also shown to reflect the different processes taking place on the networks. The distribution of influence, identified as the the number of downstream vertices, are used as a tool for investigating random vertex removal. In the transcriptional regulation networks we observe that only a small number of vertices have a large influence. The small influences of most vertices limit the effect of a random removal to, in most cases, only a small fraction of vertices in the network. The neural network has a rather different topology with respect to the influence, which are large for most vertices. To further investigate the effect of vertex removal we simulate the biological processes taking place on the networks. Opposed to the presumed large effect of random vertex removal in the neural network, the high density of edges in conjunction with the dynamics used makes the change in the state of the system to be highly localized around the removed vertex.
研究了两种不同类型的有向网络，即转录调控网络和神经网络。对有向网络结构进行了研究，结果表明其也反映了网络上发生的不同过程。将影响分布（定义为下游顶点的数量）用作研究随机顶点移除的工具。在转录调控网络中，我们观察到只有少数顶点具有较大影响。大多数顶点的影响较小，这使得在大多数情况下，随机移除的影响仅限于网络中一小部分顶点。神经网络在影响方面具有相当不同的拓扑结构，大多数顶点的影响较大。为了进一步研究顶点移除的影响，我们模拟了网络上发生的生物过程。与神经网络中随机顶点移除的假定大影响相反，高边密度与所使用的动力学相结合，使得系统状态的变化高度局限于被移除的顶点周围。
