Presently, density functional computational studies of nanostructures in heterogeneous catalysts consider either sufficiently big (''scalable with size'') unsupported metal nanoparticles (NPs) or small supported metal clusters. Both models may not be sufficiently representative of a few nm in size supported transition metal NPs dealt with in experiment. As a first step in closing the gap between theoretical models and prepared systems, we investigate the effect of a rather chemically inert oxide support, MgO(100), on relative energies and various properties of Pd and Pt NPs that consist of 49-155 atoms (1.2-1.6 nm in size) and exhibit bulk-like fcc structural arrangements. Shapes and interface configurations of metal NPs on MgO were obtained as a result of thorough optimization within the fcc motif using interatomic potentials. Then the stability and properties of the NPs were studied with a density functional method. We comprehensively characterize interaction between the NPs and MgO(100) support, their interface and effect of the support on NP properties. While the effect of MgO on relative stabilities of NPs with different shapes is found to be significant, other properties of the NPs such as electronic structure and interatomic distances within NP do not notably change upon deposition. This work paves the way to large-scale first-principles computational studies of more realistic models of oxide-supported metal catalysts. (C) 2013 AIP Publishing LLC.