Computational Methods for Transcriptional Mapping of Eukaryotic Genomes

描述(由申请人提供)：了解调控基因转录的过程对于了解进化、多细胞生物体的发展以及对包括癌症和心脏病在内的病理变化的反应至关重要。这项提议的目的是在开发和测试计算方法，然后将它们应用于实验系统方面取得实质性进展。我们开发和部署了一系列计算方法，旨在将监管者与其目标联系起来，并推断出目前未确定的监管者的目标序列。测试和验证既是针对精心策划的数据库进行的回溯性测试，也是对选定的一组预测使用各种实验方法进行的前瞻性测试。具体目标包括以下几个方面。1.开发和测试创新方法，为经过充分研究的调节器发现新的结合位点，以及为目前未确定的调节器寻找新的结合位点。前一种方法需要整合大量且往往非常庞大的数据集，然后修剪这些特征，以确定哪些是生物学上最相关的。我们的初步结果表明，这样做大大提高了现有方法的性能2.在IBM Bluegene/L上实现所有算法这是可用的最快机器之一，尽管在它上实现算法需要相当多的技术复杂性。我们目前的实施使计算能力比标准2 GHz处理器提高了约20倍。蓝色基因/L与(1)的结合使用将使研究界能够做出比目前可能的数量更多、更可靠的发现。3.在(I)S酿酒酵母全基因组和(Ii)哺乳动物GABAA受体家族上应用和检验上述方法。前者的优点是经过了充分的研究，为测试提供了大量的数据集，而且与哺乳动物基因组相比相对简单。GABA是中枢神经系统(CNS)中主要的抑制性神经递质，在中枢神经系统的发育和疾病中起着关键作用。

DESCRIPTION (provided by applicant): Understanding the processes that regulate the transcription of genes is central to understanding evolution, the development of multicellular organisms, and the response to pathological changes, including cancer and heart disease. This proposal aims to make substantial progress in developing and testing computational methods, and then applying them to experimental systems. We develop and deploy a battery of computational methods aimed at associating regulators with their targets, and inferring sequences that are targets for currently unidentified regulators. Testing and validation is carried out both retrospectively, against well curated databases, and prospectively, using a variety of experimental methods on a selected set of predictions. The specific aims include the following. 1. Develop and test innovative approaches for discovering new binding sites for well studied regulators, as well as sites for currently unidentified regulators. The former method requires integrating numerous and often very large datasets and then pruning the features to identify those that are biologically most relevant. Our preliminary results suggest that doing so substantially improves performance over existing methods 2. Implement all algorithms on IBM BlueGene/L This is one of the fastest machines available, though implementing algorithms on it requires a fair amount of technical sophistication. Our current implementation increases compute power over standard 2 GHz processors by approximately 20-fold. The use of Blue Gene/ L in combination with (1) will put the research community in a position to make discoveries that are substantially greater in number and more reliable than is currently possible. 3. Apply and test the methods on (i) the full S Cerevisiae genome and (ii) the mammalian GABA A receptor family. The former offers the advantages of being well studied, of providing a large set for data for testing, and of being relatively simple compared to the mammalian genome. GABA is the major inhibitory neurotransmitter in the central nervous system (CNS), and plays a key role in CNS development and disease.