High throughput icosahedral particle reconstruction

描述(申请人提供)：我们实验室通过结合电子冷冻显微镜中记录的数千个粒子的图像，展示了解决几个二十面体粒子的中等分辨率(7-9埃)结构的能力。在这些分辨率下，我们已经能够识别蛋白质的大部分α螺旋和β片状结构。目前，数据采集和计算机图像重建的过程是非常劳动强度的，即使对于有经验的用户来说也是如此，对于新用户来说，学习曲线非常陡峭。我们建议开发一种高通量程序，以便在3-5周内完成中分辨率二十面体粒子的结构确定过程-从数据收集到结构确定和解释。这意味着每个项目的完成时间减少到原来的1/10-1/20。这一技术进步将使分子病毒学家能够使用这种结构作为标准方法，在病毒组装以及与抗体和受体等细胞因子相互作用的背景下了解结构和功能关系。该提案由两个发展部分组成。一是在我们新获得的200kV场发射枪电子显微镜上，用一台4k×4k的ccd摄像机进行半自动在线数字图像采集。第二是提高现有软件的计算效率，并整合大部分独立软件，包括粒子拾取、显微镜参数确定、粒子定向和中心细化、三维重建和结构解释。我们将使用Python脚本语言来协调这些程序，并使用面向对象的数据库来管理整个项目数据采集和分析过程中的元数据。该软件将向研究界免费提供。除了委内瑞拉马脑炎病毒的一个新项目外，我们还将把这些技术增强部署到我们已经资助的P22噬菌体和疱疹病毒项目中，委内瑞拉马脑炎病毒是一种具有潜在生物恐怖主义威胁的精选制剂。

DESCRIPTION (provided by applicant): Our laboratory has demonstrated the capability of solving medium resolution (7-9 Angstroms) structures of several icosahedral particles by combining several thousand-particle images recorded in an electron cryomicroscope. At these resolutions, we have been able to identify most of the alpha helices and beta sheets of the proteins. Presently, the process of data collection and computer image reconstruction is very labor intensive, even for experienced users, and has a very steep learning curve for new users. We propose to develop a high throughput procedure so that the process of structure determination of icosahedral particles at medium resolution - from data collection to structure determination and interpretation - can be completed within 3-5 weeks. This represents a factor of 10-20x reduction in time for completion of each project. This technological advance will allow molecular virologists to use such structures as a standard approach to understand structure and function relationship in the context of virus assembly and interactions with cellular factors such as antibodies and receptors. This proposal consists of two developmental components. The first is to develop semi-automatic on-line digital image acquisition with a 4k x 4k CCD camera in our newly acquired 200 kV electron cryomicroscope with a field emission gun. The second is to improve the computational efficiency of existing software and to integrate much of the stand-alone software, which includes particle picking, microscope parameter determination, particle orientation and center refinement, 3D reconstruction and structural interpretation. We will use the Python scripting language to coordinate these programs and use an object-oriented database to manage the metadata throughout the project data acquisition and analysis. This software will be made freely available to the research community. We will deploy these technology enhancements to our already funded projects on P22 bacteriophage and herpes virus in addition to a new project with the Venezuelan Equine Encephalitis virus, a select agent with potential bioterrorism threat.