A simple and fast approach for missing-wedge invariant classification of subtomograms extracted from filamentous structures

Unsupervised classification of subtomograms extracted from cryo-electron tomograms is often challenging due to the presence of a missing wedge in tomographic data. Here, we propose a simple new approach to classify subtomograms extracted from cryo-electron tomograms of filamentous objects. This unsupervised classification approach uses the 1D projections of the subtomograms for classification and works independently of the orientations of the missing wedge. We applied this approach to subtomograms from eukaryotic cilia and successfully detected heterogeneity including structural polymorphism of dynein molecules.