The discovery of novel functional materials remains largely driven by tedious trial-and-error studies, whereby the rational understanding and design using modeling and simulation play an increased role thanks to more affordable computing resources. This talk surveys modeling techniques and their capabilities at the 1 to 1000 nanometer scale, describing molecular recognition mechanisms on metals, oxides, and biominerals (apatites). Applications to nanocrystal growth, catalyst design, composites, and therapeutics are shown. To support such applications, the mechanism of specific adsorption of polymers and biomacromolecules onto metallic and oxidic nanostructures will be described in atomic detail resulting from simulations with novel force fields and surface models in comparison to measurements (Figure 1). On oxidic nanoparticles such as silica and apatites, it is shown how changes in pH lead to similarity scores of attracted peptides lower than 20%, supported by model surfaces of appropriate surface chemistry and data from adsorption isotherms and crystal growth.