Computational epigenetics modeling of cell identity genes

细胞识别基因是一组功能相连的基因，它们共同实现特定细胞的表型 打字。细胞身份研究的一个主要制约因素是缺乏一种可靠的方法来定义身份目录 一种细胞类型的基因，并识别调节细胞表达网络的主要转录因子 识别基因和驱动细胞识别规范。 对我们最近的发现很感兴趣，我们假设细胞识别基因可以用表观遗传学来识别 表现出它们独特的转录调控机制的特征。我们和其他几个团体 发现细胞识别基因表现出独特的表观遗传学特征，例如广泛的H3K4me3(Chen等人， 自然遗传学，2015)和超级增强剂。我们说明了这些功能与强大的 细胞识别基因在其相关细胞类型中稳定的转录激活信号，但在其他细胞中不存在 类型。生物学家已经使用超级增强剂或广谱H3K4me3作为标记来提名细胞识别基因 最近。然而，对于大多数生物学家来说，使用这种方法仍然具有挑战性，因为它是所需的生物信息学 工具还不可用。 我们在这个提案中的总体目标是扩展我们对细胞的计算表观遗传学方法的发展 身份基因发现。利用我们的生物信息学算法DANPOS的早期成功(Chen等人， 基因组研究，2013)和DANPOS2(Chen等，自然遗传学，2015)，我们将开发一系列 (1)定义细胞识别基因的表观遗传特征，(2)为芯片序列定制参数的新算法 表观遗传学特征分析；(3)在文献检索的基础上收集已知的细胞特性基因 然后人工检查，(4)系统地鉴定未知的细胞识别基因，(5)定义主控 调节细胞识别基因网络并驱动细胞识别规范的转录因子。作为一名 原则证明，我们将应用我们的新方法来研究ECs的细胞身份决定因素 与约翰·P·库克、方龙厚和曹琦博士合作，他们是EC生物学、血管生成、 和表观遗传学。 这项研究的成功完成有望对细胞身份的研究产生广泛的积极影响 测定、转录调控和染色质表观遗传学。科学界将能够 使用本提案中开发的生物信息学工具来定义具有改进的组蛋白修饰特征 准确性，并系统地预测特定细胞类型的身份基因及其主要转录因子 在众多的生物系统或疾病模型中。我们对内皮细胞新的识别基因的功能分析将 提高对内皮细胞分化、发育和表型的机械理解，并将更好地 指导发现治疗血管疾病的治疗靶点。尽管我们关注的是组蛋白 对于EC身份基因的修饰特征，我们提出的生物信息学方法可以很容易地适应 研究所有细胞类型中的许多其他染色质标记和基因类别。

Cell identity genes are a group of functionally linked genes that jointly implement the phenotype of a given cell type. A major constraint on cell identity study is the lack of a robust method to define the catalogue of identity genes for a cell type, and to identify master transcription factors that regulate the expression network of cell identity genes and drive cell identity specification. Intrigued by our recent discoveries, we hypothesize that cell identity genes can be identified using epigenetic feature that manifests their distinct transcriptional regulation mechanism. We and several other groups discovered that cell identity genes display unique epigenetic features, e.g., broad H3K4me3 (Chen, et al, Nature Genetics, 2015) and super-enhancers. We illustrated that these features are associated with strong and stable transcription activation signals for cell identity genes in their associated cell type, but not in other cell types. Biologists have used super enhancers or broad H3K4me3 as makers to nominate cell identity genes recently. However, it is still challenging for most biologists to use this method, as the required bioinformatics tools are not yet available. Our overall goal in this proposal is to extend the development of our computational epigenetic methods for cell identity gene discovery. Leveraging the early success of our bioinformatics algorithms DANPOS (Chen, et al, Genome Research, 2013) and DANPOS2 (Chen, et al, Nature Genetics, 2015), we will develop a series of new algorithms to (1) define epigenetic features for cell identity genes, (2) customize parameters for ChIP-Seq analysis of epigenetic feature, (3) collect known cell identity genes on the basis of thorough literature search followed by manual inspection, (4) systematically identify unknown cell identity genes, and (5) define master transcription factors that regulate the network of cell identity genes and drive cell identity specification. As a proof of principle, we will apply our novel methods to study cell identity determinants for the ECs in collaboration with Drs. John P. Cooke, Longhou Fang, and Qi Cao, three experts in EC biology, angiogenesis, and epigenetics. Successful completion of this study is expected to have broad positive impact on the study of cell identity determination, transcriptional regulation, and chromatin epigenetics. The scientific community will be able to use the bioinformatics tools developed in this proposal to define histone modification features with improved accuracy, and to predict identity genes and their master transcription factors systematically for given cell types in numerous biological systems or disease models. Our functional assay for new identity genes of ECs will improve mechanistic understanding of endothelial differentiation, development, and phenotypes, and will better guide discovery of therapeutic targets for treatment of vascular diseases. Although we focus on histone modification features for EC identity genes, our proposed bioinformatics methods can be easily adapted to investigate many other chromatin marks and gene categories in all cell types.