Detecting protein structural and functional innovations from genomic data

我们的工作结合了结构生物信息学、基因组学和系统遗传学来研究蛋白质结构和功能的进化。了解蛋白质结构如何进化和编码不同的生物功能对于我们解释基因组数据以及通过合成生物学设计新功能的努力至关重要。我们方法的一个基本组成部分是从基因组数据中获得的蛋白质3D结构模型的大规模分析，用于敏感检测无法单独使用序列检测的蛋白质功能。这项工作将扩展和推广我们在关键蛋白质结构和功能适应的大规模基因组数据的计算筛选方面的工作。首先，我们打算应用这种方法从基因组数据中发现正在进行的蛋白质折叠进化的案例，这在蛋白质折叠和设计研究中具有重要的应用。其次，我们将使用这种方法在蛋白质组范围内发现谱系特异性结合位点的进化，然后对预测的相互作用进行实验表征。这项工作的一个特别重点将是鉴定植物蛋白中具有防御植物病原体作用的新的碳水化合物结合位点。第三，我们的目标是应用该方法从高多样性宏基因组序列数据中发现新的酶（如纤维素酶）变体。这将为宏基因组学辅助的蛋白质功能发现和设计创造一个框架。最终，这项工作将推进计算方法，以便从基因组数据中灵敏地检测蛋白质功能和适应性。这在生物学的一系列领域都有应用，但它也有助于我们对蛋白质多样性的基本理解，以及如何为我们自己的目的挖掘它。

Our work combines structural bioinformatics, genomics, and phylogenetics to study the evolution of protein structures and functions. Understanding how proteins structures evolve and encode different biological functions is critical in our efforts to interpret genomic data as well as design novel functions through synthetic biology.A fundamental component of our approach is large-scale analysis of protein 3D structure models derived from genomic data for sensitive detection of protein functions that cannot be detected using sequence alone. The proposed work will expand and generalize our work on computational screening of large-scale genomic data for key protein structural and functional adaptations. First, we intend to apply this approach to discover cases of ongoing protein fold evolution from genomic data, which has significant applications to protein folding and design research. Second, we will use this approach to discover evolution of lineage-specific binding sites at proteome-wide scales, followed by experimental characterization of predicted interactions. A particular focus of this work will be identification of novel carbohydrate-binding sites in plant proteins with roles in defense against plant pathogens. Third, we aim to apply the approach to discover novel enzyme (e.g., cellulase) variants from high diversity metagenomic sequence data. This will create a framework for metagenomics-aided protein function discovery and design.Ultimately, the proposed work will advance computational methods for sensitive detection of protein functions and adaptations from genomic data. This has applications in a range of areas within biology, but it also contributes to our basic understanding of protein diversity and how to mine it for our own purposes.