In October I did some docking using HADDOCK, and everything worked great. Recently, I wanted to redo some of the docks and do some new ones. Surprisingly, even with same input data I am having different results. The reason for that, is as following: I am selecting four residues as “active”, however even if I decrease the RSA limit to 5 and “Remove buried active residues from selection” is turned off, only one of them is then placed in jason config file in “activereslist” section. As result, in all clusters ligand is docked outside of the binding pocket. This one residue that is not removed from the list is most solvent accessible (what makes sense). When I did the docking in October I did not change any settings regarding RSA limits and docking worked fine - the ligand was placed right in the binding pocket close to those four “active” residues. I am using cryo-em structure and ligand present in this structure as control (I also do docking to single point mutants) so those residues I selective as active are 100% accessible and play key role i the binding, but now are removed from the active list. Were there any changes in HADDOCK algorithm since October? Are there any options I can change in order to restore solvent accessibility assessment rules to the ones were used previously?