Studying alternative splicing regulatory networks through partial correlation analysis.
Studying alternative splicing regulatory networks through partial correlation analysis.
复制标题
DOI：
10.1186/gb-2009-10-1-r3
复制
发表时间：
2009
期刊：
影响因子：
12.3
通讯作者：
Zheng S
中科院分区：
文献类型：
--
作者：
Chen L;Zheng S
在人类、小鼠和大鼠中，外显子与其调节物或靶之间以及共剪接外显子之间的连接的鉴定为可变剪接调节网络提供了新的见解。前体mRNA选择性剪接是高等真核生物蛋白质组多样性扩展的重要基因调控机制。每个剪接调节子都可能影响一大群选择性外显子。同时，每个选择性外显子受多个剪接调节子控制。高通量数据的快速积累为我们研究复杂的可变剪接调控网络提供了一个独特的机会。我们建议使用偏相关分析，以确定关联的外显子和它们的上游调控基因或其下游的靶基因之间的链接（外显子-基因链接）和共剪接外显子之间的链接（外显子-外显子链接）。偏相关分析避免了取外显子表达水平和基因表达水平这两个噪声随机变量的比值，从而获得了更高的统计功效。我们将该分析程序命名为pCastNet（剪接转录组网络的部分相关性分析）。通过对已知的替代外显子、保守模式、相对位置、功能注释和RT-PCR实验的研究，我们得出结论：pCastNet可以有效地识别外显子-基因或外显子-外显子连接。我们进一步发现，具有外显子-基因或外显子-外显子连接的基因对往往具有相似的功能或存在于相同的通路中。更有趣的是，具有外显子-基因或外显子-外显子连接的基因对倾向于共享启动子区中的顺式元件和3'非翻译区中的microRNA结合元件，这表明共选择性剪接、共转录因子结合和共microRNA结合的偶联。pCastNet构建的可变剪接调控网络有助于我们更好地理解可变剪接调控的协调性和组合性。所提出的工具可以很容易地应用于其他高通量数据，如转录组测序数据。
The identification of links between exons and their regulators or targets and between co-spliced exons in human, mouse and rat provides novel insights into the alternative splicing regulatory network. Alternative pre-mRNA splicing is an important gene regulation mechanism for expanding proteomic diversity in higher eukaryotes. Each splicing regulator can potentially influence a large group of alternative exons. Meanwhile, each alternative exon is controlled by multiple splicing regulators. The rapid accumulation of high-throughput data provides us with a unique opportunity to study the complicated alternative splicing regulatory network. We propose the use of partial correlation analysis to identify association links between exons and their upstream regulators or their downstream target genes (exon-gene links) and links between co-spliced exons (exon-exon links). The partial correlation analysis avoids taking the ratio of two noisy random variables, exon expression level and gene expression level, so that it achieves a higher statistical power. We named this analysis procedure pCastNet (partial Correlation analysis of splicing transcriptome Network). Through studies of known alternative exons, conservation patterns, relative positions, functional annotations, and RT-PCR experiments, we concluded that pCastNet can effectively identify exon-gene or exon-exon links. We further found that gene pairs with exon-gene or exon-exon links tend to have similar functions or are present in the same pathways. More interestingly, gene pairs with exon-gene or exon-exon links tend to share cis-elements in promoter regions and microRNA binding elements in 3' untranslated regions, which suggests the coupling of co-alternative-splicing, co-transcription-factor-binding, and co-microRNA-binding. Alternative splicing regulatory networks reconstructed by pCastNet can help us better understand the coordinate and combinatorial nature of alternative splicing regulation. The proposed tool can be readily applied to other high-throughput data such as transcriptome sequencing data.