I've built a 3 story RAM SS model with steel columns pinned at the base, and concrete piers (modeled as concrete columns) pinned at the top. When running a report on the member forces of each, the loads at the bottom of one do not match with the loads at the top of the other. The axial gravity loads do not match, and the axial and shear lateral loads do not match. The difference is more significant than the columns' self-weight (on the magnitude of 10 - 50 kips). I have a concrete slab-on-grade at the base to provide a diaphragm at the ground level instead of at the foundation level, however, I've detached the nodes to allow lateral loads to be carried by the framing rather than all going into the diaphragm.
Some background information on the model (all approximate):
Footprint: 200'Lx60'W with bays spread at 20' increments
Floor to floor heights: 15'-0"
Construction: Steel stick built with non-composite concrete metal deck floors and metal deck, EPDM roof
Lateral system: Mix between tension/compression chevron bracing and tension/compression X-bracing
RAM Modules used: Modeler, Beam, Column, Frame