[phenixbb] to leave atoms out of ADP restraints

Hi, all, The ADP restraints I refer to in this discussion are the ones that would make B-factors of nearby atoms to influence (or more or less resemble) each other. For some very disordered side chains, I would like to test "taking out" the "more disordered" atoms (maybe for high chi angle variability) of these restraints, for them to be more free to skyrocket, but also that they do not influence so much the main chain, in some cases where the flexibility difference between main and side chain seems to be significant. Is there a way to do so? Thanks, Jorge