I am new to the ROSIE server and Rosetta. I have found it very useful so far but am still learning. What I would like to use the server for is to design mutant variants of one protein that will provide selectivity for one binding partner over another. I have done the sequence tolerance task and I think I understand the results but now I would like to build in some predicted residues and see how they effect binding to the desired binding partner and an off-target binding partner. Can this be done on the ROSIE server? My main concern is if I dock wt-protein-A to it's receptor protein B and off-target protein C. Can I reliably compare the docking energies of these two experiments? Then is I make a mutant of Protein-A and do the same can I again reliably rank the best mutations based on the dockign energies?