We present an algorithm for determining the three-dimensional structure of molecules that possess rotational point group symmetry. Our algorithm utilizes the fact that, every two images of any such molecule share many common-lines, and that each image has several self common lines. This in turn enables a maximum-likelihood type based approach, in which all possible pairwise viewing directions are considered, thus enabling a high detection rate even in a noisy setting.

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:

Gabi Pragier (Tel Aviv University)

Keywords:

bayesian methods, computed tomography, cryo-em, image reconstruction, image registration, inverse problems, machine learning, nonlinear optimization