I want to do stochastic DCM in resting state fMRI. Should I need to include modulatory effects as assumption to get the coupling parameters (A).