In this study, we develop a theoretical framework for multiscale simulation of the formation of oriented microstructures of anisometric clay nanoparticles in polymers under shear deformations. Dissipative particle dynamics (DPD) was employed to predict the orientation process. The parameters of the DPD models were critically assessed and shown to be in agreement with previous works. We show the capability of DPD to capture the structure with far greater details than traditional orientation models such as the famous Folgar-Tucker model. Here, we propose a strategy in which the trajectories of the orientation process of weakly-interacting nanoclays were parametrized as a function of the applied shear strain instead of the time. The applied strain was used to transfer the flow filed information to the DPD models. Vice versa, it was also incorporated to pass the orientation parameters to the macroscale through a simple combination of affine and nonaffine deformations. This combination was pictured in its simplest form to be a random mixing of DPD unit cells, representing nonaffine deformations, in a larger cell which distributes an affine deformation among the unit cells. It was noted that this strategy could be used to perform multiscale simulations of orientation process provided that the unit cells represent a precise description of the interactions between the components. A comparison of this methodology with Folgar-Tucker and strain-reduction factor models proved the success of the multiscale simulation of the evolution of orientation parameters against the applied shear strain. However, it will also be discussed that the method fails to capture the microstructure evolutions, if the unit cells do not provide an accurate representation of the material, for instance in the case of strongly-interacting polymer nanocomposites