Computational Analysis of Nucleosome Positioning Signals in Eukaryotic Genomes

马萨诸塞大学医学院和华盛顿大学已经收到了一笔赠款，用于开发真核细胞内DNA包装的预测模型。这项研究将为基因表达和基因抑制的重要细胞调控机制提供重要的见解。作为第一个目标，研究小组将使用现有的基因组数据集来识别与核小体密度较高或较低相关的序列。第二个目标是设计、训练和验证基于序列的核小体定位的判别性和生成性模型。第三个目标是将训练过的模型应用于来自各种生物体和人类细胞系的数据，目的是了解DNA中编码的核小体定位信号如何与其他核小体定位因子相互作用，以及核小体定位信号的哪些方面最保守，以及在不同物种中哪些方面存在差异。该项目还将导致核小体形成的预测模型。这样的模型可以用来设计在体内形成稳定核小体的DNA序列。这些序列可以用来操纵染色质结构重塑和转录调控。在这个项目中，我们将在两个参与机构培养三名博士生。该项目将提供本科生和研究生课程的内容，以及两所大学为代表不足的少数族裔高中和本科生提供的两个暑期研究项目的内容。这个项目产生的算法、软件和数据集将通过网络免费提供给研究界。欲了解更多信息，请访问http://zlab.umassmed.edu和http://noble.gs.washington.edu/。

The University of Massachusetts Medical School and the University of Washington have received a grant to develop predictive models of the packing of DNA within eukaryotic cells. This research will provide important insights into important cellular control mechanisms for gene expression and gene suppression. As a first goal the research team will use existing genomic data sets to identify sequences that associated with higher or lower nucleosome density. The second goal is to design, train and validate sequence-based discriminative and generative models of nucleosome positioning. The third goal is to apply the trained models to data from a variety of organisms and human cell lines, with the goal of understanding how the nucleosome positioning signal encoded in the DNA interacts with other nucleosome positioning factors and which aspects of the nucleosome positioning signal are most strongly conserved, and which vary in divergent species. The project will also lead to predictive models of nucleosome formation. Such models can be used to design DNA sequences that form stable nucleosomes in vivo. These sequences can be used to manipulate chromatin structure remodeling and transcriptional regulation. In this project, we will train three PhD students at two participating institutions. The project will provide content for courses at the undergraduate and graduate levels, as well as for two summer research programs for underrepresented minority high school and undergraduate students at both institutions. The algorithms, software, and data sets resulting from this project will be made freely available to the research community via the web. Additional information may be found at http://zlab.umassmed.edu and http://noble.gs.washington.edu/