Geometry-aware template matching for cryo-electron tomograms in Dynamo

Template matching has a long history of serving as a tool for the automated analysis of volumes in cryo-electron tomography (cryo-ET). Recent theoretical and computational studies of the technique have pinpointed the importance of using fine angular samplings and high resolution scans to attain meaningful results, thus highlighting the necessity of approaches that alleviate the computational burden inherent to this technique. We present the new module for model-aware template matching in Dynamo, an open-source tool that allows the integration of possibly available a priori information in the set-up of a template matching computation, leading-in favorable cases-to large computational gains, as the scanning effort can be more efficiently restricted to relevant spatial positions and dynamically defined angular orientations. This approach has been successfully tested on a representative range of sample geometries typically arising in tomography, modeling particles distributed over tubular, membranous, or vesicular structures.