Macromolecular complexes are intrinsically flexible and often challenging to purify for structure determination by single-particle cryo-electron microscopy (cryo-EM). Such complexes can be studied by cryo-electron tomography (cryo-ET) combined with subtomogram alignment and classification, which in exceptional cases achieves subnanometer resolution, yielding insight into structure–function relationships. However, it remains challenging to apply this approach to specimens that exhibit conformational or compositional heterogeneity or are present in low abundance. To address this, we developed emClarity (https://github.com/bHimes/emClarity/wiki), a GPU-accelerated image-processing package featuring an iterative tomographic tilt-series refinement algorithm that uses subtomograms as fiducial markers and a 3D-sampling-function-compensated, multi-scale principal component analysis classification method. We demonstrate that our approach offers substantial improvement in the resolution of maps and in the separation of different functional states of macromolecular complexes compared with current state-of-the-art software.

大分子复合物具有内在的灵活性，通常难以通过单颗粒低温电子显微镜（cryo-EM）进行结构鉴定。可以通过冷冻电子断层扫描（cryo-ET）结合子层析图对准和分类来研究此类复合物，在特殊情况下，可以实现亚纳米分辨率，从而深入了解结构与功能之间的关系。然而，将这种方法应用于表现出构象或组成异质性或以低丰度存在的标本仍然具有挑战性。为了解决这个问题，我们开发了emClarity（https://github.com/bHimes/emClarity/wiki），这是GPU加速的图像处理程序包，具有迭代层析X射线倾斜系列优化算法，该算法使用子图作为基准标记和3D-采样功能补偿的 多尺度主成分分析分类方法。我们证明，与当前最先进的软件相比，我们的方法在地图的分辨率和大分子复合物不同功能状态的分离方面提供了实质性的改进。