Is it possible to run phenix.autosol MIR problem with two "native" datasets? And does this make sense, or would it be better to scale the two natives together to make a single dataset before running autosol? eab