Understanding conformational dynamics from macromolecular crystal diffuse scattering

All macromolecular crystals contain some extent of disorder. The diffraction from such crystals contains diffuse scattering in addition to Bragg peaks and this scattering contains information about correlated displacements in the constituent molecules. While much work has been performed recently in decoding the dynamics of the crystalline ordering, the goal of understanding the internal dynamics of the molecules within a unit cell has been out-of-reach. In this article, we propose a general framework to extract the internal conformational modes of a macromolecule from diffuse scattering data. We combine insights on the distribution of diffuse scattering from short- and long-range disorder with a Bayesian global optimization algorithm to obtain the best fitting internal motion modes to the data. To illustrate the efficacy of the method, we apply it to a publicly available dataset from triclinic lysozyme. Our mostly parameter-free approach can enable the recovery of a much richer, dynamic structure from macromolecular crystallography.