Integrative Analysis of Public RNA-seq Repositories

描述（由申请人提供）：最近，RNA-seq技术越来越多地取代微阵列用于表达谱分析。在本提案中，我们将及时应对快速积累的RNA-seq数据带来的新挑战和机遇。我们将设计新的方法对多个RNA-seq数据集进行整合分析，以研究选择性剪接的功能和调控。具体而言，我们有以下具体目标：（1）我们将开发一种新的基于图的模式挖掘方法，以重建剪接模块的图谱，并确定相关的实验条件，在人类，小鼠，苍蝇和酵母。(2)我们将通过利用RNA-seq数据提供的表达和剪接信息来研究转录和剪接这两个重要的调控过程之间的耦合。我们将设计一种新的多层网络挖掘方法来系统地识别耦合的转录剪接模块。(3)我们将预测选择性剪接的功能， 转录本，以建立高分辨率的人类基因组功能注释。预测的功能将被纳入GeneOntology和Ensembl数据库，以造福生物界。(4)我们将对目标1、2、3中的计算预测子集进行实验验证。(5)我们将开发网络数据库和软件，以直接使科学界受益。我们的方法和软件将大大促进重复使用 大量现有的RNA-seq数据，减少产生新数据的必要性，并提高我们对各种干扰下基因调控的理解。

DESCRIPTION (provided by applicant): Recently, the RNA-seq technology is increasingly replacing microarray for expression profiling. In this proposal, we will timely address the emerging challenges and opportunities brought by the rapidly accumulating RNA-seq data. We will design novel methods to perform integrative analysis of many RNA-seq datasets to study the functions and regulations of alternative splicing. In particular, we have the following specifi aims: (1) We will develop a novel graph- based pattern mining method to reconstruct an atlas of splicing modules and identify the associated experimental conditions in human, mouse, fly, and yeast. (2) We will study the coupling between transcription and splicing, the two important regulatory processes, by exploiting both expression and splicing information provided by RNA-seq data. We will design a novel multi-layer network mining approach to systematically identify coupled transcription- splicing modules. (3) We will predict the functions of alternatively spliced transcripts to establish a high-resolution function annotation of human genome. The predicted functions will be incorporated into the GeneOntology and the Ensembl databases to benefit the biological community. (4) We will perform experimental validation on a subset of computational predictions made in Aims 1, 2, 3. (5) We will develop web databases and software to directly benefit the scientific community. Our methods and software will significantly facilitate the re-use of the vast amount of existing RNA-seq data, reduce the necessity to generate new data, and improve our understanding of gene regulations under a variety of perturbations.