Modeling the structure & evolution of regulatory regions in eukaryotic gemomes

描述(申请人提供)：转录是基因表达调控的核心，然而转录调控的计算分析目前面临着许多挑战和机遇：大量测序的基因组允许研究和利用调控序列的保守性，但在严格框架下这样做的算法仍然很少。时空基因表达的详细数据已经变得可用，使我们能够利用这些信息来阐明复杂生物体发育过程中的调控相互作用。长期目标是建立计算模型，以推断调控网络及其在模式生物乃至最终人类发展过程中的演变。这一特定建议的目标是开发算法来分析序列水平上的基因调控的保守性，以及一种综合的方法来对对发育重要的保守的调控区域进行建模。其具体目的是：(1)基于对两个苍蝇基因组中核心启动子元件的保守性的比较研究，破译确定功能转录起始点的准确要求，并建立全基因组比较注释的模型。(2)开发和实现一种基于系统发育隐马尔可夫模型的非编码调控序列递进多重比对算法，并研究核心启动子在更广泛的物种集合中的进化。(3)将该算法建立的框架扩展到更复杂的调控模块(如发育增强子和E2F靶基因)，并纳入关于假定的上游因素的先验信息，以预测调控相互作用。计算预测将通过少量的实验得到验证。这项拟议的研究有望促进对调控区域进化的理解，以及如何建立准确利用几个物种的序列信息的计算模型。与公众健康相关：了解基因调控是如何在基因组中编码的，无疑是当今分子生物学中最有趣的挑战之一，而且直觉上认为，这种机制中发生的错误会导致基因的错误表达，而且在基于基因的疾病中往往很重要。我们的研究将有助于从计算和实验上找到DNA中准确的调节区，并了解控制模式生物和人类基因表达的机制。

DESCRIPTION (provided by applicant): Transcription is at the heart of the regulation of gene expression, yet the computational analysis of transcription regulation currently faces a number of challenges and opportunities: The large number of sequenced genomes allows to study and exploit the conservation of regulatory sequences, but algorithms that do so in a rigorous framework are still scarce. Detailed data of spatiotemporal gene expression has become available, enabling us to use this information to elucidate regulatory interactions in the development of complex organisms. The long-term goal is to build computational models to infer regulatory networks and their evolution in the development of model organisms and ultimately humans. The objective of this particular proposal is to develop algorithms to analyze the conservation of gene regulation on the sequence level, as well as an integrated approach to model conserved regulatory regions important for development. Its specific aims are: (1) To decipher the precise requirements to define a functional transcription start site, based on a comparative study of the conservation of core promoter elements in two fly genomes, and build a model for genome-wide comparative annotation. (2) To develop and implement an efficient progressive multiple alignment algorithm for non-coding regulatory sequences based on phylogenetic hidden Markov models, and to study the evolution of core promoters in a wider set of species. (3) To extend the framework set by this algorithm to more complex regulatory modules (such as developmental enhancers and E2F target genes), and to incorporate prior information on putative upstream factors to predict regulatory interactions. Computational predictions will be validated by a small number of experiments. The proposed research is expected to advance the understanding on the evolution of regulatory regions, and how to build computational models that accurately utilize sequence information from several species. Relevance to public health: Understanding how gene regulation is encoded in the genome is undoubtedly one of the most interesting challenges in molecular biology today, and it is intuitive that errors occurring in this machinery lead to mis-expression of genes, and may often be important in genetically based diseases. Our research will help to find the exact regulatory regions in DNA, both computationally and experimentally, and to learn the mechanisms that control the expression of genes in model organisms and humans.