基于球坐标系的冷冻电镜二十面体病毒高分辨三维重构的算法研究

项目名称： 基于球坐标系的冷冻电镜二十面体病毒高分辨三维重构的算法研究

项目编号： No.91230116

项目类型： 重大研究计划

立项/批准年度： 2013

项目学科： 生物物理、生化与生物分子学、生物力学与组织工程

项目作者： 刘红荣

作者单位： 湖南师范大学

项目金额： 70万元

中文摘要： 病毒三维结构的研究是了解病毒致病机制、疾病预防及药物设计的基础。病毒颗粒通常体积巨大、结构复杂、难于长成晶体，传统的X射线晶体学和核磁共振技术难以精确求解其结构。近年来，随着电子显微技术的发展、算法改进、计算能力的提高，冷冻电镜单颗粒技术已成为破解这一难题的最有力手段。由于采用低剂量电子成像，冷冻电镜照片的噪声非常大，要从数以万计的二维投影照片恢复高分辨三维结构，重构算法是关键。基于球坐标系的生物大分子三维重构方法由项目申请人及所在课题组最先提出，该算法在抑制噪声、提高分辨率取得了一定成效。本项目拟通过改进冷冻电镜图像处理的算法和程序，并应用于具体的生物病毒三维重构，进一步提高病毒重构分辨率，推动冷冻电镜技术向前发展。研究内容主要包括：1)基于球坐标系的高精度病毒三维重构的快速算法研究；2)基于球坐标系的MPI并行程序设计；3)具体的病毒三维结构研究。最终形成一套快速、稳定的并行重构程序。

中文关键词： 冷冻电镜；二十面体病毒；三维重构；对称失配；算法研究

英文摘要： Three-dimensional (3D) structural studies of viruses are crucial to understanding their pathogenic mechanism, disease prevention and drug design. Generally, viruses are very large and complex, and are difficult to grow into crystal, so it’s not easy to obtain their high-resolution 3D structures by the traditional methods -- X-ray crystallography and nuclear magnetic resonance spectroscopy (NMR). With the recent development of microscopy, data-processing algorithm and program, and computing power, the cryo-electron microscopy and single-particle reconstruction (commonly referred to collectively as “cryo-EM”) has become the most powerful tool for analyses the high-resolution structures of viruses. Due to the low-dose imaging, the cryo-EM images are very noise (the signal-to-noise ratio less than 0.1), and the reconstruction algorithm is the key to restore the virus structures from tens of thousands particle images. The new reconstruction algorithm based on the spherical coordinate system and icosahedral symmetry-adapted functions (ISAFs) was firstly put forward by the project applicants and their groups, and this method has obtained some results on noise suppression and resolution improvement. The major idea of this project is to enhance the resolution of 3D reconstruction of virus by improving the algorithm

英文关键词： Cryo-electrom microscopy；Icosahedral virus；Three-dimensional reconstruction；Symmetry-mismatch reconstruction；Algorithm research