Virtual Talk: High throughput, high performance algorithms for high resolution subtomogram averaging

Mikhail Kudryashev
Stiftung Max-Delbrück-Centrum für Molekulare Medizin (MDC)

Subtomogram averaging from cryo electron tomograms in a powerful method to determine structures of macromolecules in their native state. Outstanding applications to protein lattices, coats and ribosomes provided unique insights into their functions and even revealed interactions with small molecules in situ. For other macromolecules, such as membrane proteins, which are present in tomograms in limited numbers, the throughput of data processing and the processing time are key bottlenecks in obtaining high-resolution reconstructions. I will introduce the tools that we developed in our lab for in situ structural biology. TomoBEAR is a workflow for processing of tomographic data utilizing common cryo-EM tools and original code that allows transparent near-automated tomographic pre-processing, alignment, reconstruction and particle identification. Our substack-based analysis using SUSAN implemented for high performance computing allows up to several orders of magnitude speedup and produces high-resolution maps. I will demonstrate the applications of our tools to structure of an ion channel that we determined up to side-chain resolution.

Back to Long Programs