冷冻电镜已经成为了一项应用越来越广泛的生物物理学技术。特别是近些年来，由于最新的信号探测器的应用，以及单颗粒三维重建算法的革新和不断改进，冷冻电镜技术取得了飞跃式的发展。通过该技术解析得到的生物大分子结构可以在单分子条件下达到近原子分辨率（<4Å）水平。随着越来越多具有重要生物医学功能的蛋白质及其复合体的高分辨结构被解析出来，冷冻电镜技术正在逐渐引发一场分子生物学的技术革命，并引起全世界的关注。细胞中大多数蛋白的功能都是通过复杂的构象变化来实现的。因此，要全面认识蛋白质结构和功能关系，就不能不研究蛋白质的动力学行为和活性动态构象的变化规律。本项目将集中力量从数据分析算法上入手，着眼于发展对冷冻电镜单颗粒图像数据的多构象三维分类、动态结构和高效并行的数学物理计算方法，从而将冷冻电镜技术发展到一个新的高度。

Cryo-electron microscopy (Cryo-EM) has become a widely used biophysical technology. In recent years, with the application of the direct electron detector device camera, as well as the innovation and continuous improvement of the single-particle reconstruction algorithms, cryo-EM has made leaps and bounds. The macromolecular structures determined by single particle cryo-EM can reach near atomic resolution (<4Å). With more and more high-resolution proteins or complexes with important biomedical implications determined, cryo-EM is gradually leading to a revolution in structural biology and attracting the worldwide attention. Most protein complexes in cells use complicated conformational transition to fulfill their biological functions. Thus, to fully understand the mechanistic relationship between protein structure and function, one has to study their structural dynamics and the foundation of conformational changes. This proposal seeks to investigate novel algorithms for cryo-EM data analysis, focusing on the problems of conformational 3D classification, reconstructions of dynamic structures and highly efficient parallel computing methods. The proposed studies are expected to push the envelope of the cryo-EM methodology.