Smoothness

Statistics

Bayesian probability

Algorithm

Machine learning

Mathematical analysis

Artificial intelligence

Estimation theory

Expectation–maximization algorithm

Mathematics

Computer science

Maximum likelihood

Kriging

Likelihood function

Maximum a posteriori estimation

IUCrJ

A new method to estimate the trajectories of particle motion and the amount of cumulative beam damage in electron cryo-microscopy (cryo-EM) single-particle analysis is presented. The motion within the sample is modelled through the use of Gaussian process regression. This allows a prior likelihood that favours spatially and temporally smooth motion to be associated with each hypothetical set of particle trajectories without imposing hard constraints. This formulation enables the a posteriori likelihood of a set of particle trajectories to be expressed as a product of that prior likelihood and an observation likelihood given by the data, and this a posteriori likelihood to then be maximized. Since the smoothness prior requires three parameters that describe the statistics of the observed motion, an efficient stochastic method to estimate these parameters is also proposed. Finally, a practical algorithm is proposed that estimates the average amount of cumulative radiation damage as a function of radiation dose and spatial frequency, and then fits relative B factors to that damage in a robust way. The method is evaluated on three publicly available data sets, and its usefulness is illustrated by comparison with state-of-the-art methods and previously published results. The new method has been implemented as Bayesian polishing in RELION -3, where it replaces the existing particle-polishing method, as it outperforms the latter in all tests conducted.

A Bayesian approach to beam-induced motion correction in cryo-EM single-particle analysis