Computational Modeling of Mammalian Promoters

描述(由申请者提供)：长期目标是了解人类基因转录是如何控制和调节的。这个假设是，这样的理解可以通过开发数学模型来实现，这些模型利用当地的遗传和表观遗传信息来预测启动子位置和组织特异性活性。最近，大规模的实验技术已经在基因组中定位了大量的活性启动子，虽然功能强大，但它们的假阳性(由于异常的，可能是无功能的mRNA转录本)、假阴性(由于组织和发育阶段的不完全采样)和其他错误(由于方案偏差)仍然不确定。因此，重要的是有更多的方法结合更全面或更严格的标准，并检查序列特征，除了阐明分子机制外，还可以允许计算预测和直接实验检测更多的启动子。即使绘制了所有人类启动子的图谱，仅仅记录它们的位置也不能告诉我们它们是如何被识别和部署进行转录的。因此，随着更多的实验图谱数据的获得，开发数学模型来理解启动子的结构、功能和进化变得更加重要。现在，随着人类基因组的完整测序和几乎所有蛋白质编码基因的定位，了解这些基因是如何控制和调控的已经成为基因组研究中的一大挑战。由于一个基因通常可以通过在不同细胞、不同发育阶段和/或对不同信号的反应而产生多个转录本，在ENCODE计划的推动下，在构建更全面的基因调控网络之前，了解定义和调节替代启动子的关键元件将是一项关键任务，用于解决此类问题的新的高通量基因组学技术正在快速发展。先进的计算方法与实验验证相结合，对于最终理解基因表达的调控机制是必不可少的。新的具体目标是：A1。提取、比较和分类哺乳动物中的组织特异性启动子，以便它们可以被分成不同的(不一定是相互排斥的)表达和/或表观遗传学类别；确定顺式调控基序/模块作为启动子结构特征及其与组织特异性染色质和表达模式的关系；建立组织特异性启动子和表达预测的数学模型；在选定的组织中进行真实调控途径的案例研究。这项拟议的研究将结合实验和计算方法和技术，以便更好地了解哺乳动物启动子的遗传和表观遗传顺式调控密码。这样的模型可能会为基因调控或错误调控的机制提供新的见解，并将为进一步的基因表达全球调控的功能研究产生许多假设。

DESCRIPTION (provided by applicant): The long-term goal is to understand how human gene transcription is controlled and regulated. The hypothesis is that such an understanding may be achieved by developing mathematical models that are predictive of promoter position and tissue-specific activity by using local genetic and epigenetic information. Recently, large- scale experimental technologies have mapped a great number of active promoters in a genome and while powerful, their rates of false positives (due to aberrant, likely nonfunctional mRNA transcripts), false negatives (due to incomplete sampling of tissues and developmental stages), and other errors (due to protocol biases) remain uncertain. Consequently, it is important to have additional approaches that incorporate more comprehensive or stringent criteria, and to examine sequence characteristics that, in addition to illuminating molecular mechanisms, may permit computational prediction and direct experimental detection of additional promoters. Even when all human promoters are mapped, merely documenting their positions will not tell us how they are recognized and deployed for transcription. Therefore, as more experimental mapping data become available, the more essential it becomes to develop mathematical models to understand promoter architecture, function and evolution. Now with the complete sequencing of the human genome and localization of almost all of the protein coding genes, understanding how each of these genes are controlled and regulated has become a major challenge in the genome research. Since a gene can often produce multiple transcripts through alternative promoter usage in different cells, at different developmental stages and/or in response to different signals, understanding key elements that define and regulate alternative promoters will be a crucial task before more comprehensive gene regulation networks can be constructed Powered by the ENCODE project, new high throughput genomics technologies for attacking such problems are being developed at a rapid pace. Advanced computational approaches coupled with experimental validations are essential for the ultimate understanding of the regulatory mechanisms of gene expression. The new specific aims are: A1. Extract, compare and classify tissue-specific promoters in mammals so that they may be grouped into different (not necessarily mutually exclusive) expressional and/or epigenetical classes; A2. Identify cis-regulatory motifs/modules as promoter architecture features and their relation to tissue-specific chromatin and expression patterns; A3. Build mathematical models for tissue-specific promoter and expression predictions; A4. Conduct case studies in real regulation pathways in selected tissues. The proposed research will combine experimental and computational approaches and technologies in order to better understand mammalian promoters in terms of genetic and epigenetic cis-regulatory codes. Such models are likely to offer new insights into mechanisms of gene regulation or mis-regulation, and will generate many hypotheses for further functional studies on global regulation of gene expression.