High-Resolution Modeling of Protein-Protein Interaction Networks Within and Between Species

宿主-微生物系统生物学新兴领域的一个主要焦点是阐明宿主-微生物蛋白质-蛋白质相互作用网络的一般原理。然而，目前的分析对宿主-微生物相互作用的洞察有限，因为它们在高度抽象的水平上操作，将蛋白质视为节点，将蛋白质-蛋白质相互作用视为边缘。一个关键的瓶颈是缺乏在基因组水平上对宿主-微生物生物分子相互作用网络的三维(3D)原子观点。最近，我们构建了第一张人蛋白和病毒蛋白之间相互作用网络的3D结构图。由此产生的宿主-病毒结构相互作用网络对宿主-病毒相互作用提供了重要的见解。然而，我们初步绘制的人-微生物结构相互作用网络图在覆盖率和准确性方面存在严重限制。 拟议研究计划的长期目标是开发和应用新方法，以尽可能高的准确度和覆盖率为已知的人-人和人-微生物蛋白质-蛋白质相互作用建立3D原子结构模型。这种细胞电路的高分辨率视图将使算法开发能够更好地预测基因功能，更好地理解正常和疾病生物学，能够发现控制物种内和物种间相互作用的新原理，否则这些原理就隐藏在二元网络中，并能够构建用于预测性模拟细胞行为的计算机全细胞模型。 为了实现这一长期目标，提出了五个短期目标。首先，我们将使用同源建模来构建人内和人-微生物蛋白质-蛋白质相互作用的3D结构模型。其次，我们将提高人内和人-微生物结构相互作用网络的覆盖率，不仅为球形结构域之间的相互作用建立结构模型，而且为线性基序结合域和包含线性多肽基序的底物结构域之间同样丰富的相互作用建立结构模型。第三，通过开发基于线程和对接的结构相互作用建模新算法，提高人内和人与微生物结构相互作用网络的覆盖率。第四，我们将通过显式使用3D结构信息来改进序列比对，从而建立准确的人内和人-微生物蛋白质-蛋白质相互作用的结构模型。第五，我们将通过计算分析和实验合作相结合的方式来验证我们的预测。 拟议的研究计划将通过构建原本隐藏在二元网络中的人-人和人-微生物相互作用的3D结构视图，显著推进宿主-微生物系统生物学的最先进水平。这项拟议的工作是首次将结构系统生物学方法应用于宿主-微生物相互作用和传染病研究。该研究计划的成功完成将带来人-人和人-微生物蛋白质-蛋白质相互作用的结构模型，具有显著提高的准确性和覆盖率，为宿主内和宿主-微生物蛋白质-蛋白质相互作用网络的结构建模提供新的工具，并提高对宿主-微生物相互作用和传染病的系统水平的理解。最后，这项提议的执行将提供一套算法、工具和数据集，用于最大限度地发挥结构系统生物学方法对宿主-微生物相互作用研究的影响。

A major focus of the emerging field of host-microbe systems biology is the elucidation of general principles governing host-microbe protein-protein interaction networks. However, current analyses generate limited insights into host-microbe interactions as they operate on a highly abstract level, treating proteins as nodes and protein-protein interactions as edges. A key bottleneck is the lack of a three-dimensional (3D), atomistic view of the host-microbe biomolecular interaction network at the genomic scale. Recently, we constructed the first 3D structural map of the network of interactions among human proteins and viral proteins. The resulting host-virus structural interaction network provided significant insights into host-virus interactions. However, our preliminary map of human-microbe structural interaction network is severely limited in coverage and accuracy. The long-term objective of the proposed research program is to develop and apply novel methods to build 3D atomistic structural models for known human-human and human-microbe protein-protein interactions with the highest possible accuracy and coverage. Such a high-resolution view of cellular circuitry will enable algorithm development for better prediction of gene function and better understanding of normal and disease biology, enable the discovery of new principles governing intra- and inter-species interactions that are otherwise hidden in the binary network, and enable the construction of in silico whole-cell models for predictive simulation of cell behavior. To achieve this long-term objective, five short-term objectives are proposed. First, we will construct 3D structural models for within-human and human-microbe protein-protein interactions using homology modeling. Second, we will improve the coverage of within-human and human-microbe structural interaction networks by building structural models not only for interactions between globular domains, but also for the equally abundant interactions between linear motif binding domains and the substrate domains that contain the linear peptide motifs. Third, we will improve the coverage of within-human and human-microbe structural interaction networks by developing new threading- and docking-based algorithms for structural interaction modeling. Fourth, we will build accurate structural models of within-human and human-microbe protein-protein interactions by explicitly using 3D structural information to improve sequence alignment. Fifth, we will validate our predictions by a combination of computational analyses and experimental collaborations. The proposed research program will significantly advance the state-of-the-art in host-microbe systems biology by constructing a 3D structural view of human-human and human-microbe interactions that are otherwise hidden in the binary network. The proposed work is the first to apply a structural systems biology approach to host-microbe interaction and infectious disease research. Successful completion of the research program will lead to structural models of human-human and human-microbe protein-protein interactions with significantly improved accuracy and coverage, new tools for structural modeling of within-host and host-microbe protein-protein interaction networks, and improved systems-level understanding of host-microbe interactions and infectious diseases. Finally, the execution of this proposal will provide a set of algorithms, tools, and datasets that serve to maximize the impact of structural systems biology approaches on host-microbe interaction research.