Structural ensemble refinement with cryo-EM

Abstract

Cryo-electron microscopy (cryo-EM) is a state-of-the-art experimental technique for structural characterization of biomolecules at near-atomic resolution. In Cryo-EM, thousands to millions of noisy, randomly-oriented 2D images of the biomolecule of interest are obtained, and these images are typically used to reconstruct 3D density volumes of the biomolecule. This reconstruction process is challenging due to the low signal-to-noise ratio of the images, and also heterogeneity in the structural ensemble of the biomolecule. To resolve this problem, we propose a framework that introduces a prior set of conformations obtained, for example from molecular simulations, to retrieve the structural ensemble captured by the cryo-EM images.