Integrative modeling core

核心5：集成建模 核心摘要 我们的目标是使项目和其他核心能够执行基于结构的发现和优化配体 病毒蛋白靶标。这一目标将通过开发和应用我们独特的综合模型来实现 工具箱，计算目标病毒系统的结构模型，基于来自其他系统的各种实验数据 核心和项目。获得这些模型的关键挑战是病毒蛋白的结构异质性， 包括二级结构内容和结构域取向的大变化以及 环状和侧链构象。准确、准确、完整地描述和描述这些 多状态是理解和调节其与配体功能的关键。我们假设，明确的 尤其需要对病毒蛋白的多种构象进行建模：病毒蛋白质组可能已经进化到 更多地利用构象的多样性来传递功能，而不是利用不受限制的蛋白质组 到少量的短蛋白质。我们将通过两种方式应对这一挑战。第一，粗粒度结构 模型(目标1)，基于冷冻电子显微镜、冷冻电子断层扫描、 化学交联和足迹技术将为原子结构提供必要的步骤；例如，通过 适用于X射线结晶学的变种设计和高分辨率单粒子冷冻启动模型的设计 电磁重建。第二，明确描述其异质性的病毒蛋白质的多态原子模型 将根据X射线结晶学(目标2)、低温电子显微镜(目标3)的数据进行计算，以及 配基构效关系研究(目标4)。我们的多态模型将有助于揭示静态、 动态的，甚至是神秘的病毒蛋白结合口袋，这反过来将促进配体的发现和 优化。所有的核心和项目都将依赖于我们的模型。

CORE 5: INTEGRATIVE MODELING CORE SUMMARY We aim to enable Projects and other Cores to perform structure-based discovery and optimization of ligands for viral protein targets. This goal will be achieved by developing and applying our unique integrative modeling toolbox to compute structural models of the target viral systems, based on varied experimental data from other Cores and Projects. The key challenge in obtaining these models is structural heterogeneity of viral proteins, including large variations in secondary structure content and domain orientations as well as small variations in loop and side chain conformations. Accurate, precise, and complete description and characterization of these multiple states is key to understanding and modulating their functions with ligands. We hypothesize that explicit modeling of multiple conformations of viral proteins is especially needed: Viral proteomes may have evolved to exploit the multiplicity of conformations for delivering function more than the proteomes that are not constrained to a small number of short proteins. We will address this challenge in two ways. First, coarse-grained structural models (Aim 1), based on limited information from cryo-electron microscopy, cryo-electron tomography, chemical cross-linking, and footprinting, will provide an essential step to atomic structures; for example, via design of variants suitable for X-ray crystallography and starting models for high-resolution single particle cryo- EM reconstruction. Second, multi-state atomic models of viral proteins that explicitly describe their heterogeneity will be computed based on data from X-ray crystallography (Aim 2), cryo-electron microscopy (Aim 3), and ligand structure-activity relationship (SAR) studies (Aim 4). Our multi-state models will help reveal static, dynamic, and even cryptic binding pockets of viral proteins, which, in turn, will facilitate ligand discovery and optimization. All Cores and Projects will rely on our models.