Understanding dynamics of biological macromolecular complexes by estimating a mechanical model via statistical mechanics from cryo electron microscopy images

Abstract

Cryo electron microscopy (cryo EM) imaging experiments can lead to stochastic models for biological macromolecular complexes. However, interpreting the statistical variability is difficult. In some situations, the variability in the original complexes is due primarily to thermal fluctuations which are snap frozen in place during the preparation of the specimen. In this case the images are images of samples of the equilibrium statistical mechanics ensemble of the complex. Based on representing the complex by a spring-mass mechanical model, an estimation problem for determining the masses and spring constants is described and demonstrated on synthetic data. With a model, quantities such as normal modes can be computed, which provide insight into the dynamics of biological complexes.