Development of a Database for the Discovery and Annotation of Functional Sites in Protein Structures

本提案旨在开发新颖和通用的计算工具来发现蛋白质结构中的功能位点。这些位点几乎控制着细胞化学的所有方面，它们的鉴定具有重要的应用，例如了解蛋白质功能的基础，以及通过工程蛋白质模拟物或抑制剂来修饰细胞通路。然而，到目前为止，功能位点只能在彻底的突变分析之后才能可靠地确定：这是一个缓慢、劳动密集型的实验室过程，受检测灵敏度的限制，并且是蛋白质特异性的。在这里，我们建议用新的算法来识别功能位点，用于序列分析和蛋白质结构中重要特征的有效几何比较。这建立在我们对进化痕迹（ET）方法的初步工作的基础上。首先，在具有足够进化信息的蛋白质中，我们将使用ET识别活性位点的关键几何和化学特征，并将其总结为三维模型。其次，在缺乏足够进化信息的蛋白质结构中，我们将寻找模仿这些霉菌的相似区域。当大量的原始序列和结构数据压倒了传统的分析方法时，这种低成本的策略将揭示与功能最相关的蛋白质结构区域，因此与实验最相关。

This proposal aims to develop novel and general computational tools to discover functional sites in protein structures. These sites control nearly all aspects of cellular chemistry and their identification has important applications, such as to understand the basis of protein function, and to modify cellular pathways by engineering protein mimetics or inhibitors. Until now, however, functional sites could be identified reliably only after exhaustive mutational analysis: a laboratory process that is slow, labor intensive, limited by assay sensitivity, and which is protein specific. Here, we propose instead to identify functional sites with new algorithms for sequence analysis and for the efficient geometric comparison of important features in protein structures. This builds on our preliminary work on the Evolutionary Trace (ET) method. First, in proteins with sufficient evolutionary information, we will use ET to identify the key geometric and chemical features of active sites and summarize these into 3-dimensional molds. Second, in protein structures lacking sufficient evolutionary information, we will search for look-alike areas that mimic these molds. At a time when the outpour of raw sequence and structure date overwhelms conventional means of analysis, this low-cost strategy will reveal the regions of protein structures that are most relevant to function, and therefore to experimentation.