面向典型疾病的高精度生物大分子三维结构研究

冷冻电镜图像三维重构是研究生物大分子三维结构的前沿技术,其产生的大规模、大尺度图像数据需要高精度、高效率的算法进行处理。但是目前冷冻电镜三维重构技术存在如下缺陷：1)缺乏面向大规模冷冻电镜图像的高精度单颗粒重构算法；2)缺乏面向大尺度冷冻电镜图像的高效电子断层三维重构算法；3)缺乏针对生物大分子冷冻电镜三维重构可视化分析工具；4)缺乏处理大规模、大尺度的生物大分子冷冻电镜数据的高效并行处理技术。针对上述问题,本项目将从五个方面对冷冻电镜图像三维重构开展研究：1)研究面向大规模的冷冻电镜图像单颗粒重构算法；2)研究面向大尺度的冷冻电镜图像电子断层重构算法；3)研究能够对T级冷冻电镜重构结果可视化分析技术；4)研究基于“天河二号”的大规模、大尺度生物大分子冷冻电镜数据的并行处理技术；5）本项目将提供一套完整的针对大规模、大尺度冷冻电镜图像的三维重构系统，为生物学家提供强有力的软件工具。

Three-dimensional (3D) reconstruction of cryo-electron microscopy imaging has been a leading technique in analyzing 3D structures of biomacromolecules. However, current techniques for 3D reconstruction of cryo-electron microscopic imaging, and the corresponding visualization techniques, have the limitations in both their processing speed and the accuracy of the reconstructed structures. As larger and larger amount of biological data produced from cryo-electron microscopy, techniques and systems that can efficiently process T-level cryo-electron microscopic data, including data preprocessing, 3-D reconstruction, and real-time visualization, are urgently demanded. This project will conduct research on the following topics: (1) Algorithms for single particle analysis and electron tomography 3-D reconstruction with more efficiency and more accuracy. (2) Efficient visualization techniques for T-level data of 3-D reconstruction. (3) Tianhe-2 supercomputer system structure oriented parallel processing techniques for large scale cryo-electron microscopic data. (4) Construction of a software system on Tianhe-2 system for large scale cryoelectronic microscopy data processing, 3-D reconstruction, and visualization. Our objective is to provide efficient techniques and convenient software tools for scientists for their research on biomacromoleclar structures.