IEEE/ACM Trans Comput Biol Bioinform. 2018 Nov-Dec;15(6):2074-2078. doi: 10.1109/TCBB.2018.2796572. Epub 2018 Jan 23.
The numerous interconnections within complex systems enable us to control networks towards a desired state through a few suitable selected nodes, which are called driver nodes. Recent works analyzed directed human Protein-Protein Interaction (PPI) network based on structural control theory. They found that indispensable proteins, whose removal increase the number of driver nodes, are the primary targets of human viruses and drugs. However, the human PPI network is usually incomplete and may include many false-positive or false-negative interactions. That prompts us to ask whether these indispensable proteins are stable to possible structural changes. Here, we present a method to alter the type of indispensable proteins and thereby investigate the stability of indispensable proteins. By comparing the sets of indispensable proteins before and after structural changes to the network, we find that very few added or removed interactions can change the type of many indispensable nodes. Furthermore, some indispensable proteins are very sensitive to structural changes and have significantly lower interactions than the other indispensable proteins. The results indicate that indispensable proteins are sensitive to structural changes. Therefore, approaches based on structural control theory should be used with caution because of the incomplete nature of these networks.
复杂系统中的众多相互关系使我们能够通过少数几个合适的选择节点（称为驱动节点）将网络控制到期望的状态。最近的研究基于结构控制理论分析了有向人类蛋白质-蛋白质相互作用（PPI）网络。他们发现，不可缺少的蛋白质（去除这些蛋白质会增加驱动节点的数量）是人类病毒和药物的主要靶点。然而，人类 PPI 网络通常是不完整的，并且可能包含许多假阳性或假阴性相互作用。这促使我们提出这样一个问题：这些不可缺少的蛋白质是否能够抵抗可能发生的结构变化。在这里，我们提出了一种改变不可缺少蛋白质类型的方法，从而研究不可缺少蛋白质的稳定性。通过比较结构变化前后网络中不可缺少蛋白质的集合，我们发现很少有添加或删除的相互作用可以改变许多不可缺少节点的类型。此外，一些不可缺少的蛋白质对结构变化非常敏感，其相互作用明显低于其他不可缺少的蛋白质。结果表明，不可缺少的蛋白质对结构变化很敏感。因此，由于这些网络的不完整性，基于结构控制理论的方法应该谨慎使用。
