High-throughput cryo-ET structural pattern mining by unsupervised deep iterative subtomogram clustering

• Published in: Proceedings of the National Academy of Sciences - PNAS 2023-04, Vol.120 (15), p.e2213149120-e2213149120
• Main Authors: Zeng, Xiangrui, Kahng, Anson, Xue, Liang, Mahamid, Julia, Chang, Yi-Wei, Xu, Min
• Format: Article
• Language: English
• Subjects: Biological Sciences; Cellular structure; Cluster Analysis; Clustering; Cryoelectron Microscopy - methods; Data mining; Deep learning; Electron Microscope Tomography - methods; Homogeneous structure; Image Processing, Computer-Assisted - methods; Iterative methods; Macromolecular Substances - chemistry; Macromolecules; Molecular Structure; Pattern analysis; Physical Sciences; Three dimensional models
• ISSN: 0027-8424; 1091-6490
• DOI: 10.1073/pnas.2213149120

Cryoelectron tomography directly visualizes heterogeneous macromolecular structures in their native and complex cellular environments. However, existing computer-assisted structure sorting approaches are low throughput or inherently limited due to their dependency on available templates and manual labels. Here, we introduce a high-throughput template-and-label-free deep learning approach, Deep Iterative Subtomogram Clustering Approach (DISCA), that automatically detects subsets of homogeneous structures by learning and modeling 3D structural features and their distributions. Evaluation on five experimental cryo-ET datasets shows that an unsupervised deep learning based method can detect diverse structures with a wide range of molecular sizes. This unsupervised detection paves the way for systematic unbiased recognition of macromolecular complexes in situ.