Discovery of Protein Seq.Struct.Func.Relationships

描述（由申请人提供）：拟议研究的目的是开发和提供一个用户友好的计算平台，用于研究蛋白质序列-结构-功能关系。计算工具将被开发和系统地评估，以利用不同信息（序列特征、结构特征、基因表达数据、蛋白质相互作用数据等）的各种组合，目的是能够使用它们从多个角度研究功能。这个发现的计算平台可以显著地推进我们对分子水平生物学机制的理解，增强基因注释能力，并可能导致新的治疗方法。该项目的具体目标包括：开发一套算法（包括旨在克服现有方法局限性的新算法）和用于将蛋白质序列分配给结构或功能家族的数据表示；软件的快速和灵活组装数据集，从多个异质蛋白质数据存储库，以支持蛋白质序列-结构-功能关系的分析；用于获取蛋白质功能重要部分的序列和结构相关性的计算工具；描述和预测蛋白质相互作用的计算工具；以及一套可扩展的用于分析蛋白质结构和功能的软件模块。使用这些工具产生的预测将直接在特定生物学问题的背景下进行测试-预测受体激酶的候选配体，预测参与新型SH3-SH2相互作用的残基，以及抗体可变区域的组装。算法、软件、数据和文档将免费提供。将灵活的数据集生成工具与分析蛋白质序列-结构-功能关系的软件模块集成在一起，是对目前情况的巨大改进。由此产生的计算平台和工具将在结构生物学、信号转导和功能基因组学等领域找到大量用户。这项研究将与爱荷华州立大学计算机科学、生物信息学和计算生物学研究生的研究培训紧密结合。

DESCRIPTION (provided by applicant): The aim of the proposed research is to develop and provide a user-friendly computational platform for investigating protein sequence-structure-function relationships. Computational tools will be developed and systematically evaluated to exploit various combinations of diverse information (sequence features, structural features, gene expression data, protein interaction data, etc.) with the aim of being able to use them to investigate function from multiple viewpoints. This computational platform for discovery can significantly advance our understanding of biological mechanisms at the molecular level, enhance gene annotation capabilities, and potentially lead to new therapeutic methods. Specific aims of the project include the development of: A suite of algorithms (including new algorithms designed to overcome limitations of existing approaches) and data representations for assigning protein sequences to structural or functional families; Software for rapid and flexible assembly of data sets derived from multiple heterogeneous protein data repositories to support analysis of protein sequence-structure-function relationships; Computational tools for eliciting sequence and structural correlates of functionally important parts of proteins; Computational tools for characterizing and predicting protein-protein interactions; and A set of extensible software modules for analysis of protein structure and function. Predictions generated using these tools will be tested directly in the context of specific biological problems - prediction of candidate ligands for a receptor kinases and prediction of residues that participate in a novel SH3-SH2 interaction, and assembly of antibody variable regions. The algorithms, software, data, and documentation will be made freely available. The proposed integration of flexible tools for generating data sets with software modules for analyzing protein sequence-structure-function relationships represents an enormous improvement over the present situation. The resulting computational platform and tools will find a large community of users in areas ranging from structural biology, to signal transduction, to functional genomics. This research will be closely integrated into the research-based training of graduate students in Computer Science and Bioinformatics and Computational Biology at Iowa State University.