We present a coarse-grained model for using dissipative particle dynamics simulations to study the interaction between cis-1,4 polybutadiene polymer chains and a metal oxide (Cu2O) surface at the mesoscopic scale. We employ a bottom-up approach in order to link the structural properties with their underlying molecular properties over a wide range of time and length scales. The parameterization of the model was realized using a recently developed Bayesian scheme that is based on trajectory matching of an atomistic description of the system. The model is used to investigate the local structure of polymers in a confined slab geometry by means of density profiles, radius of gyration, orientation with respect to the surface, and their adsorption.