The functioning of a single cell or organism is governed by the laws of chemistry and physics. The bridge from biology to chemistry and physics is provided by structural biology: to understand the functioning of a cell, it is necessary to know the atomic structure of macromolecular assemblies, which may contain hundreds of components. To characterise the structures of the increasingly large and often flexible complexes, high resolution structure determination (as was possible for example for the ribosome) will likely stay the exception, and multiple sources of structural data at multiple resolutions are employed. Integrating these data into one consistent picture poses particular difficulties, since data are much more sparse than in high resolution methods, and the data sets from heterogeneous sources are of highly different and unknown quality and may be mutually inconsistent, and that data are in general averaged over large ensembles and long times. Molecular modelling, a crucial element of any structure determination, plays an even more important role in these multi-scale and multi-technique approaches, not only to obtain structures from the data, but also to evaluate their reliability. This proposal is to develop a consistent framework for this highly complex data integration problem, principally based on Bayesian probability theory. Appropriate models for the major types data types used in hybrid approaches will be developed, as well as representations to include structural knowledge for the components of the complexes, at multiple scales. The new methods will be applied to a series of problems with increasing complexity, going from the determination of protein complexes with high resolution information, over low resolution structures based on protein-protein interaction data such as the nuclear pore, to the genome organisation in the nucleus.