Landscape of transcription in human and mouse

该项目的总体目标是使用各种高通量测序平台在人类和小鼠（C57 BL/6 NJ）组织和原代细胞系中生成精细结构RNA图谱，通过确定是否可以识别其翻译产物的证据来评估新转录物的生物学重要性。从分析的每个样品中，我们建议以生物学重复分离长（>200个核苷酸）和短（< 200个核苷酸）RNA。这些样品的基于Illumina的图谱最初将使用（1）核糖体（r-）RNA耗尽的长总RNA的RNA测序（-seq）生成。(2)烟草酸性焦磷酸酶（TAP）预处理的短RNA的RNA-seq（3）总RNA的基因表达的对端帽分析（PE-CAGE）。此外，对于原代细胞系的子集，我们将从细胞核和细胞质亚细胞级分产生上述文库。长RNA-Seq数据将被提炼成由剪接点、多腺苷酸化位点和从头基因和转录物组成的功能元件。短RNA数据将被提取到代表短RNA的5'端的重叠群中，直到读取长度。将分析PE-CAGE数据以形成代表与转录本主体内部的标签连接的转录本的5'末端的簇。重要的是，将使用非参数不可重现检测率（nplDR）脚本评估每个元素的重现性。总的来说，这些数据将允许检测 新的转录区域和支持信息的启动子区域的位置和转录本的亚细胞驻留。总之，这些数据将被用来生成模型的非编码和蛋白质编码转录本，并区分异构体在复杂的基因座，以获得全面的哺乳动物转录组。对于这些样本的一个子集，我们将同时收集人类供体的基因组序列，以提供一个参考图谱，该图谱将用于绘制RNA数据，并获得有关等位基因特异性表达和RNA编辑的信息。将使用长距离（PacBio/454）测序测试未注释的转录模型。最后，将进行蛋白基因组学分析，并将结果与未注释的转录本进行比较。

The overall goal of this project is to generate fine-structure RNA maps in human and mouse (C57BL/6NJ) tissues and primary cell lines using a variety of high-throughput sequencing platforms, to evaluate the biological importance of novel transcripts by determining if evidence of their translated products can be identified. From each sample analyzed, we propose to isolate long (>200 nucleotides) and short (< 200 nucleotides) RNA in biological duplicate. Illumina-based maps for these samples will initially be generated using (1) RNA sequencing (-seq) of ribosomal (r-)RNA depleted long total RNA. (2) RNA-seq of tobacco acid pyrophosphatase (TAP) pre-treated short RNA (3) Pair-end Cap Analysis of Gene Expression (PE-CAGE) of total RNA. Additionally, for a subset of primary cell lines we will generate the above libraries from nuclear and cytoplasmic subcellular fractions. Long RNA-Seq data will be distilled down into functional elements consisting of splice junctions, polyadenylatio sites and de novo genes and transcripts. The short RNA data will be distilled into contigs representing the 5' ends of short RNAs up to the read length. PE-CAGE data will be analyzed to form clusters representing the 5' ends of transcripts linked to a tag internal to the transcript body. Importantly, each element will be assessed for reproducibility using a nonparametric Irreproducible Detection Rate (nplDR) script. Collectively, these data will allow for the detection of novel transcribed regions and supportive information as to the location of promoter regions and subcellular residence of transcripts. In aggregate, these data will be used to generate models of both noncoding and protein coding transcripts and to distinguish isoforms at complex loci necessary to obtain a comprehensive view of mammalian transcriptomes. For a subset of these samples we will simultaneously collect the genome sequence of the human donors to provide a reference map that will be used to map the RNA data against and derive information concerning allele-specific expression and RNA editing. Unannotated transcript models will be tested using long-range (PacBio/454) sequencing. Lastly, proteogenomic analysis will be done and the results compared against the unannotated transcripts.