An integrative genomic strategy to infer global RNA regulatory networks

描述（由申请人提供）：越来越多的人认识到，哺乳动物的生物复杂性通过RNA复杂性的调节而极大地放大。这种调节是由数百种RNA结合蛋白（RBP）和microRNA（miRNA）通过与其靶标的序列特异性相互作用介导的。RNA的错误调节可导致许多遗传疾病。尽管他们的关键作用，努力研究RNA的复杂性在很大程度上阻碍了准确推断全球RNA调控网络的困难，由于高通量的实验技术和有效的计算方法的缺陷。本申请描述了实验和计算方法的组合，以促进在系统水平上对哺乳动物大脑中的体内RNA调节的理解。在实验方面，我将利用洛克菲勒大学罗伯特·达内尔实验室建立的小鼠遗传系统和高通量技术，指导阶段的研究将在那里进行。HITS-CLIP将用于生成几种重要神经元RBP的生化足迹的全基因组图谱;外显子连接微阵列和RNA-Seq将用于生成核苷酸解析转录组图谱，用于野生型大脑和缺乏个体RBP的大脑以及发育中的小鼠大脑的比较分析。在计算方面，将采用综合建模技术，如隐马尔可夫模型（HALF）和贝叶斯网络，对来自多个数据源的生物化学、结构、基因组和进化信息进行概率建模，以便开发RBP/miRNA靶位点和RNA调控网络的高度预测模型。从Nova对RNA剪接调控的分析中获得了大量的初步数据，这证明了这种整合基因组策略在定义准确和全面的RNA调控网络并获得新的生物学见解方面的有效性。在这里，我建议进一步改进这些方法，并将该策略扩展到其他RBP，单独或组合，以及RNA调控的其他步骤。在基因工程系统中发现的调控机制将进一步扩展到研究发育中的大脑和不同脑区的组合和动态RNA调控。虽然我在机器学习和RNA调控的计算研究方面接受了广泛的培训，但这个职业发展奖将使我能够扩展现有的技能，并继续发展我的实验技能。在达内尔实验室和洛克菲勒大学的优良环境将极大地促进不仅指导研究，而且我的过渡到一个独立的学术地位。总之，这项研究将为我未来的研究铺平道路，这些研究旨在解码正常生物过程和人类疾病中RNA调控的规则。 公共卫生相关性：RNA调控对于哺乳动物基因表达、生物体复杂性和人类疾病的多样化至关重要。该研究将结合高通量实验技术和新的综合计算方法，在系统水平上研究神经元RNA复杂性的基本原理。

DESCRIPTION (provided by applicant): It is increasingly recognized that mammalian biologic complexity is amplified enormously by the regulation of RNA complexity. This regulation is mediated by hundreds of RNA-binding proteins (RBPs) and microRNAs (miRNAs) through sequence-specific interactions with their targets. Misregulation of RNAs can cause a number of genetic diseases. Despite their critical roles, efforts to study RNA complexity are largely impeded by the difficulty to accurately infer global RNA-regulatory networks, due to deficiencies of high-throughput experimental technologies and effective computational methods. This application describes a combination of experimental and computational approaches to advance the understanding of in vivo RNA regulation in mammalian brains at the systems level. On the experimental side, I will take advantage of the mouse genetic systems and high-throughput technologies established in the Robert Darnell laboratory at Rockefeller University, where the mentored phase of research will be performed. HITS-CLIP will be used to generate genome-wide maps of biochemical footprints of several important neuronal RBPs; exon-junction microarrays and RNA-Seq will be used to generate nucleotide-resolution transcriptome profiles for comparative analysis of wild type brains and brains lacking individual RBPs, and of developing mouse brains. On the computational side, integrative modeling techniques, such as hidden Markov models (HMMs) and Bayesian networks, will be employed to probabilistically model biochemical, structural, genomic, and evolutionary information from multiple data sources, so that highly predictive models of RBP/miRNA target sites and RNA-regulatory networks can be developed. Substantial preliminary data have been obtained from the analysis of RNA splicing regulation by Nova, which demonstrates the effectiveness of such an integrative genomic strategy to define accurate and comprehensive RNA-regulatory networks and obtain novel biological insights. Here I propose to further improve these methods, and extend the strategy to other RBPs, individually or in combination, and other steps of RNA regulation. The regulatory mechanisms discovered in genetically engineered systems will be further extended to study the combinatorial and dynamic RNA regulation in developing brains and in different brain regions. While I have received extensive training in machine learning and computational studies of RNA regulation, this career development award will allow me to expand my existing skills and continue to develop my experimental skills. The excellent environment in the Darnell lab and Rockefeller University will greatly facilitate not only the mentored research, but also my transition to an independent academic position. Together, the proposed study will pave the road to launch my future investigations that aim to decode rules governing RNA regulation in normal biological processes and human diseases. PUBLIC HEALTH RELEVANCE: RNA regulation is critical for the diversification of mammalian gene expression, organism complexity, and human diseases. The proposed study is to investigate the fundamentals of neuronal RNA complexity at the systems level, using a combination of high-throughput experimental technologies and novel integrative computational methods.