Issues with residues and structure changing with FlexPepDock; is this normal?

This is probably a silly question so I apologize if so, but recently lab has been investigating some novel interaction sites between a protein and peptide based off of recent crosslinking and IP data. I created some pdb files of the protein and the peptide aligned at the various novel and established crosslinking sites and submitted them to FlexPepDock to see if the models were reasonable.

At first I was excited because the rmsBB scores were very good, but when my PI and I opened the top 10 pdbs sent back to us we discovered that not only had the shape of the proteins changed but many of the residues were different than the ones in the pdb file I submited; I even reopened the one I had sbumitted to double check that I had not accidently changed the residues when I had saved it. I'm not sure why it happened, and we are unsure if we can use the results because of this. Is this a normal thing that happens with FlexPepDock, and if not, is there a way to fix it? No one in our lab is really familiar with the software at all.