Icosahedral Particle Reconstruction Software

描述（由申请人提供）：许多球形病毒和一些酶复合物形成具有二十面体对称的颗粒。电子低温显微镜和单粒子重建的使用使我们能够将这类粒子的三维结构可视化到7-9 A的分辨率。这个分辨率足以分辨长螺旋和识别大薄片。在过去的三年中，我们在这个分辨率范围内确定了6种尺寸从600 -1250 A的二十面体颗粒的结构。它们包括疱疹病毒衣壳、水稻矮病毒、原衣壳和成熟形式的P22噬菌体、轮状病毒和细胞质多角体病毒。有了这些经验，我们已经确定了针对二十面体粒子重建的软件套件中的一些可能的改进，以提高其未来维护和升级的便利性，确定结构的准确性，生物学最终用户的易用性和计算效率。在这项提议中，我们将有四个具体目标。第一个目标是用c++重写主要代码，其中包括二十面体粒子中心和取向的细化以及傅里叶贝塞尔重构。它们将针对共享和分布式计算机平台进行优化，并将以模块化形式编写，使它们易于维护和修改。二是改进粒子取向参数细化算法、对比度传递函数和B因子校正算法以及傅里叶贝塞尔重构算法。我们期望这些改进的编码将导致更精确的二十面体粒子的结构测定。第三个目标是将改进的代码与Python语言绑定器以及用户友好的图形界面集成在一起，并输出相关的中间数据处理结果，以允许用户决定改进是否朝着正确的结构收敛方向发展。第四个目标是通过帮助菜单充分记录代码，并通过互联网广泛和自由地传播它们。在软件发布之前，我们将通过模拟和实验数据验证其可靠性和准确性。实验数据由我们之前确定的亚纳米分辨率的结构组成，其中一些成分的晶体结构是已知的。

DESCRIPTION (provided by applicant): Many spherical viruses and some enzyme complexes form particles having icosahedral symmetry. The use of electron cryomicroscopy and single particle reconstruction has allowed us to visualize the 3-dimensional structures of this type of particle to 7-9 A resolution. This resolution is sufficient to resolve long alpha helices and recognize large beta sheets. In the past three years, we have determined 6 structures of icosahedral particles with sizes ranging from 600 -1250 A in this resolution range. They include herpesvirus capsid, rice dwarf virus, procapsid and mature forms of P22 bacteriophage, rotavirus and cytoplasmic polyhedrosis virus. With this experience, we have identified a number of possible improvements in the suite of software specific to the icosahedral particle reconstruction in order to improve the ease of its future maintenance and upgrade, the accuracy of determined structure, the ease of usage by biology end-users and the computational efficiency. In this proposal, we will have four specific goals. The first aim is to rewrite the major codes in C++, which includes the icosahedral particle center and orientation refinement and the Fourier Bessel reconstruction. They will be optimized for both shared and distributed computer platforms and will be written in modular form to make them easy to maintain and modify. The second aim is to improve the algorithms for particle orientation parameter refinement, contrast transfer function and B factor corrections and Fourier Bessel reconstruction respectively. We expect that these improved codes will lead to a more accurate structure determination of icosahedral particles. The third aim is to integrate the improved codes with a Python language binder together with a user-friendly graphical interface and to output relevant intermediate data processing results that would allow the users to decide if the refinement is heading to the correct structure convergence. The fourth aim is to document the codes adequately with a help menu and to disseminate them broadly and freely through the internet access. Before the software release, we will confirm their reliability and accuracy by testing with simulated and experimental data. The experimental data consist of structures that we have previously determined to subnanometer resolution and for which the crystal structures of some of their components are known.