Dissipative Particle Dynamics (DPD) simulations were conducted to study the rheological properties of linear polymers undergoing steady and unsteady shear flows. The interaction parameters were derived from all-atom molecular dynamics simulations through the concept of potential of mean force. Although, the coarse-graining approach used correctly predicts the structural properties of the polymeric systems at equilibrium, the softness of the interactions leads to unrealistically fast dynamics. In order to minimize this effect, the DPD model accounts with a dissipative term responsible to slow down the particles mobility, but a correct value of its strength is still difficult to evaluate on a multiscale basis. Here, the time-scale of the mesoscopic simulations is corrected by matching the diffusivity (through the mean-squared displacement) between coarse-grained and atomistic models at equilibrium. This procedure provides very good estimation of most of the dynamical properties relevant to rheology, turning the friction coefficient a rather arbitrary choice. Hence, a time-mapping function can be created relating the friction coefficient with the corresponding time scaling corrections. The consistency between the equilibrium properties, such as bulk viscosity and longest relaxation times, and the flow curves obtained from non-equilibrium simulations, proofs the validity of the scalings performed. Keywords: Dissipative particle dynamics, rheology, time-mapping, coarse-graining