Factor-general characterization of dynamic transcriptional stress responses

描述（由申请人提供）： 自从十年前人类基因组首次测序以来，研究人员在确定多种功能元件的基因组位置方面取得了很大进展，包括控制基因调控的序列。然而，迄今为止的主要焦点是对单个监管要素进行分类，而不考虑它们的动态行为或相互作用。在这项提案中，我们概述了一种创新的方法，用于识别对基因调控至关重要的序列，并表征它们的动态相互作用。我们的建议涉及将一种直接测量基因表达的强大方法（称为PRO-seq）与DNase-seq的适应相结合，DNase-seq是一种用于识别基因调控转录因子结合的基因组位置的方法。我们建议在诱导系统刺激后的时间过程中应用这些方法，以获得关于结合和表达的动态的全基因组信息，特别关注由小分子雷公藤红素在永生化K562白血病细胞系中诱导的应激反应。因为PRO-seq和DNase-seq都不依赖于特定转录因子的抗体，也不依赖于染色质免疫沉淀技术，我们将这种方法描述为通用因子和ChIP。我们的建议有三个主要目标：（1）使用PRO-seq识别和表征转录单位;（2）使用DNase-seq识别和表征许多转录因子的结合位点;（3）整合这些转录和结合的动态模式，以揭示调控序列和转录单位之间的相互作用网络。这些目标中的每一个都涉及新的统计模型和计算方法的开发。我们新生成的数据、我们的预测和我们的软件都将公开。

DESCRIPTION (provided by applicant): Since the human genome was first sequenced a decade ago, researchers have made great strides in identifying the genomic locations of many kinds of functional elements, including the sequences that control gene regulation. Nevertheless, the primary focus to date has been to catalog individual regulatory elements, without regard for their dynamic behavior or interactions. In this proposal, we outline an innovative approach for both identifying sequences critical for gene regulation and characterizing their dynamic interactions. Our proposal involves combining a powerful method for directly measuring the expression of genes, called PRO-seq, with an adaptation of DNase-seq, a method for identifying positions in the genome at which gene-regulating transcription factors are bound. We propose to apply these methods in a time course after stimulation of an inducible system to obtain dynamic, genome-wide information about both binding and expression, focusing in particular on stress responses induced by the small molecular celastrol in the immortalized K562 leukemia cell line. Because neither PRO-seq nor DNase-seq depends on antibodies to particular transcription factors, or on the technique of chromatin immunoprecipitation, we describe this approach as factor-general and ChIP-free. Our proposal has three main aims: (1) to identify and characterize transcription units using PRO-seq; (2) to identify and characterize the binding sites for many transcription factors using DNase-seq; and (3) to integrate these dynamic patterns of transcription and binding to reveal networks of interaction between regulatory sequences and transcription units. Each of these aims involves the development of new statistical models and computational methods. Our newly generated data, our predictions, and our software will all be made publicly available.