A single biological molecule typically exhibits multiple structural configurations, each with a different function. Cryo-electron microscopy can capture this variability since different structures yield different projection images. We discuss methods for characterizing this variability which exploit the fact that it is often well-approximated by a low-rank model. This property allows us to draw on methods from matrix completion and singular value shrinkage. We demonstrate the utility and efficiency of the proposed methods on experimental datasets.

This presentation is part of Minisymposium “MS51 - Algorithms for Single Particle Reconstruction in Cryo-Electron Microscopy (cryo-EM). (3 parts)”
organized by: Roy Lederman (Yale University) , Joakim Andén (Flatiron Institute) .

Authors:

Joakim Andén (Flatiron Institute)

Keywords:

computed tomography, cryo-electron microscopy, image reconstruction, inverse problems, low-rank, nonlinear optimization, numerical linear algebra, single-particle analysis, volume data