We aim to develop and apply computational methods for integrative determination of the structures and dynamics of macromolecular assemblies, in turn informing the function and evolution of these systems as well as how to modulate them. The broad goal is to contribute to a predictive spatiotemporal model of the cell. This goal is being achieved by a formal integration of experiment, physics, and statistical inference, spanning all relevant size and time scales. Our computational methods are implemented in the open source Integrative Modeling Platform package (IMP) and the resulting models are deposited in the PDB-Dev database. This research enhances the discovery of general principles that underlie all cellular processes, which in turn also facilitates drug discovery.