Chapter 8 - Cryo-electron microscopy

Abstract

Cryo-electron microscopy (cryo-EM) is based on the principle of imaging radiosensitive objects in a transmission electron microscope under cryogenic conditions. Electrons, accelerated toward the sample by electrical potential differences, interact with the macromolecules embedded in the sample. As a result of the interaction, some of the incident electrons are scattered. Thereafter, the electrons are focused to produce an image by the electromagnetic lenses of the microscope. Images are generated by phase contrast caused by the interference between the unscattered and scattered parts of the exit wave. Single-particle cryo-EM works with enormously large datasets and averages similar images prior to further processing and interpretation. To compress the data and make the process computationally efficient, multivariate statistical analysis and eigenvector–eigenvalue-based principal component analysis are carried out. The analyses involve computation of the distance between images in the data cloud, followed by classification in the abridged eigenvector space. Statistical image processing seeks to find the model of a macromolecule that has the maximum likelihood of being the correct one.