Subtomogram averaging allows for high-resolution structure determination in near-native environments.

While cryo-ET allows for direct 3D visualization of a specimen, tomograms typically have very low signal-to-noise ratios and poor resolution. This is a basic limiting factor, as biological specimens are very sensitive to electron-induced damage. Subtomogram averaging, a 3D analog to single-particle analysis, allows for high-resolution structure determination using cryo-ET data. Any repeating structures, including single molecules or subunits within macromolecular assemblies, can be averaged to improve resolution.

Using subtomogram averaging, we aim to determine high-resolution structures of viral proteins within intact viruses and virus-like particles. Under optimal conditions, near-atomic resolutions are achievable by subtomogram averaging, but we will also use complimentary methods when necessary, including single particle cryo-EM, X-ray crystallography, and molecular modeling. However, subtomogram averaging remains our core method, as its broad resolution range and direct visualization allows us to tie together structural data with near-native biological context.

Subtomogram averaging is a developing method, and part of our focus is to develop and optimize computational methods. Much of this work feeds into STOPGAP, our open-source MATLAB-based subtomogram averaging package.