Academic

Single-particle cryo-electron microscopy (cryo-EM) is a tool which aids in the analysis of proteins and biological macromolecules, to investigate their molecular, biochemical and cellular processes. The biological sample is vitrified, allowing the molecules to be imaged in its near-native state. To reduce radiation damage, many molecules are imaged at a low signal-to-noise ratio, at random and unknown orientations.

At the heart of cryo-EM is attaining a 3-dimensional (3D) reconstruction from 2-dimensional (2D) images. Existing reconstruction methods are reviewed, beginning with classical approaches, followed by several key approaches widely used today. However, these existing methods produce only a single 3D structure, or at most a discrete set of 3D structures. Of current interest is the continuous heterogeneity problem, which remains largely unsolved. To address this problem, this thesis proposes a probabilistic framework. By placing distributions on the unknown orientations and 3D structure, it allows for uncertainty in the reconstruction to be modelled in a robust way. This proposed method is then investigated on simulated and real-world datasets.