Living organisms rely on the specific recognition of groups of molecules in practically every biological process. Hence, the importance of understanding molecular recognition and determining the structures of molecular complexes cannot be overestimated. Here we propose the development of a new approach that combines intermediate resolution data of complexes from electron microscopy with atomic-resolution structures of the complex components in order to obtain reliable atomic models of the complex in a fast and automatic fashion. The proposed approach combines correlation-statistics based docking and real-space refinement of the atomic structures into the density derived by electron microscopy with tools from theory based protein-protein docking, flexible structure alignment, and comparative modeling. We will drive the development through applications to actin-talin complexes and bacterial pili. By coupling the development of the new system with applications that address fundamental question in cell biology and biomedicine, we will obtain immediate feedback as to how our approach performs in practice while providing valuable insights through the application of these advanced tools. ? ? ?