Degree dependence in rates of transcription factor evolution explains the unusual structure of transcription networks
Degree dependence in rates of transcription factor evolution explains the unusual structure of transcription networks
复制标题
DOI：
10.1098/rspb.2009.0210
复制
发表时间：
2009-07-07
影响因子：
4.7
通讯作者：
Pomiankowski, Andrew
中科院分区：
文献类型：
--
作者：
Stewart, Alexander J.;Seymour, Robert M.;Pomiankowski, Andrew
转录网络有一个不寻常的结构。在原核生物和真核生物中，每个转录因子所调控的靶基因的数量，即其出度，都遵循宽尾分布。相比之下，调节靶基因的转录因子的数量，其入度，遵循一个窄得多的分布，没有宽尾。我们构建了一个转录网络进化的模型，通过反式和顺式突变，基因复制和删除。这些不同的进化过程对网络结构的影响足以产生不对称的入度和出度分布。然而，复制已知的入度和出度分布所需的参数值是不现实的。然后，我们考虑了基因进化速率的变化取决于其在网络中的位置。当具有许多调节相互作用的转录因子被限制为比具有很少相互作用的转录因子进化得更慢时，转录网络的入度和出度分布的细节可以在一系列合理的参数值上完全再现。我们的模型产生的网络依赖于不同进化过程的相对速率。通过确定具有正确度分布的网络产生的环境，我们能够评估模型中不同进化过程在进化过程中的相对重要性。
Transcription networks have an unusual structure. In both prokaryotes and eukaryotes, the number of target genes regulated by each transcription factor, its out-degree, follows a broad tailed distribution. By contrast, the number of transcription factors regulating a target gene, its in-degree, follows a much narrower distribution, which has no broad tail. We constructed a model of transcription network evolution through trans- and cis-mutations, gene duplication and deletion. The effects of these different evolutionary processes on the network structure are enough to produce an asymmetrical in- and out-degree distribution. However, the parameter values required to replicate known in- and out-degree distributions are unrealistic. We then considered variation in the rate of evolution of a gene dependent upon its position in the network. When transcription factors with many regulatory interactions are constrained to evolve more slowly than those with few interactions, the details of the in- and out-degree distributions of transcription networks can be fully reproduced over a range of plausible parameter values. The networks produced by our model depend on the relative rates of the different evolutionary processes. By determining the circumstances under which the networks with the correct degree distributions are produced, we are able to assess the relative importance of the different evolutionary processes in our model during evolution.