Jiang W, Li Z, Zhang Z, Booth C R, Baker M L, Chiu W
Program in Structural and Computational Biology and Molecular Biophysics, Baylor College of Medicine, Houston, Texas 77030, USA.
J Struct Biol. 2001 Dec;136(3):214-25. doi: 10.1006/jsbi.2002.4439.
Electron cryomicroscopy of large macromolecular complexes is becoming an increasingly powerful tool for revealing three-dimensional structures without the need for crystallization. The execution of image processing, however, requires experience and is error-prone due to the need for a human operator to carry out interactive and repetitive processes. We have designed an approach which is intended to make image processing simple and rapid, both for experts and for novice users. We demonstrate this approach using the well-established reconstruction scheme for icosahedral particles. Finally, we implement semi-automated virus reconstruction (SAVR), an expert system that integrates the most CPU intensive and iterative steps using the scripting language Python. SAVR is portable across platforms and has been parallelized to run on both shared and distributed memory platforms. SAVR also allows the incorporation of new algorithms and facilitates the management of the increasingly large data sets needed to achieve higher resolution reconstructions. The package has been successfully applied to several data sets and shown capable of generating icosahedral reconstructions to sub-nanometer resolutions (7-10 A ).
对大型大分子复合物进行电子冷冻显微镜成像正日益成为一种强大的工具，可用于在无需结晶的情况下揭示三维结构。然而，图像处理的执行需要经验，并且由于需要人工操作员进行交互式和重复性过程，因此容易出错。我们设计了一种方法，旨在使图像处理对于专家和新手用户而言都简单且快速。我们使用针对二十面体颗粒的成熟重建方案来演示此方法。最后，我们实现了半自动病毒重建（SAVR），这是一个专家系统，它使用脚本语言Python集成了最耗费CPU且具有迭代性的步骤。SAVR可跨平台移植，并且已并行化以在共享内存平台和分布式内存平台上运行。SAVR还允许纳入新算法，并有助于管理为实现更高分辨率重建所需的越来越大的数据集。该软件包已成功应用于多个数据集，并显示能够生成分辨率达到亚纳米级（7 - 10埃）的二十面体重建结果。
