Studying the functional conservation of cis-regulatory modules and their transcriptional output
Studying the functional conservation of cis-regulatory modules and their transcriptional output
复制标题
DOI：
10.1186/1471-2105-9-220
复制
发表时间：
2008-04-29
影响因子：
3
通讯作者：
Bailey, Timothy L.
中科院分区：
文献类型：
--
作者：
Bauer, Denis C.;Bailey, Timothy L.
背景资料：顺式调控模块（CRMs）是围绕靶基因的独特的基因组区域，其可以独立地激活启动子以驱动转录。CRM的激活受转录因子（TF）的某种组合的结合控制。如果能够预测特定CRM介导的转录输出，这将是非常有益的。同样有益的是通过计算机识别特定的CRM作为特定组织或情况下表达的驱动因素。我们扩展了最近开发的生化建模方法来管理这两个预测任务。给定一组TF，它们的蛋白质浓度，以及每个TF在假定CRM中的位置和结合强度，该模型预测基因的转录输出。我们的方法预测的调节CRM的位置，通过使用预测的TF结合位点的基因附近的区域作为输入到模型和搜索的区域，产生一个预测的转录率最接近匹配的已知率。结果：在这里，我们显示的能力的CRM调节的前夕基因，MSE 2的例子上的模型。在D.黑腹果蝇的基因被应用于其他七种果蝇的eve基因的周围序列。该模型成功地预测了8种果蝇中6种果蝇的MSE2定位和输出。黑腹果蝇与其他果蝇物种的关系，并准确预测MSE2在这些物种中的位置和转录输出。然而，我们还表明，目前的模型不足以作为全基因组CRM扫描仪，因为它错误地预测其他基因组区域是MSE 2。
Background: Cis-regulatory modules (CRMs) are distinct, genomic regions surrounding the target gene that can independently activate the promoter to drive transcription. The activation of a CRM is controlled by the binding of a certain combination of transcription factors (TFs). It would be of great benefit if the transcriptional output mediated by a specific CRM could be predicted. Of equal benefit would be identifying in silico a specific CRM as the driver of the expression in a specific tissue or situation. We extend a recently developed biochemical modeling approach to manage both prediction tasks. Given a set of TFs, their protein concentrations, and the positions and binding strengths of each of the TFs in a putative CRM, the model predicts the transcriptional output of the gene. Our approach predicts the location of the regulating CRM by using predicted TF binding sites in regions near the gene as input to the model and searching for the region that yields a predicted transcription rate most closely matching the known rate.Results: Here we show the ability of the model on the example of one of the CRMs regulating the eve gene, MSE2. A model trained on the MSE2 in D. melanogaster was applied to the surrounding sequence of the eve gene in seven other Drosophila species. The model successfully predicts the correct MSE2 location and output in six out of eight Drosophila species we examine.Conclusion: The model is able to generalize from D. melanogaster to other Drosophila species and accurately predicts the location and transcriptional output of MSE2 in those species. However, we also show that the current model is not specific enough to function as a genome-wide CRM scanner, because it incorrectly predicts other genomic regions to be MSE2s.