Hi, I'm trying to refine a structure with a disulfide bond between a cysteine and a thiol-containing peptide. There are two conformations for the thiopeptide, in one of them the thiol is facing away from SG of the cysteine (~7 A away). I defined the link as: refinement.pdb_interpretation { apply_cif_link { data_link = DRG-CYS residue_selection_1 = chain A and resname DRG and resid 10 residue_selection_2 = chain B and resname CYS and resid 20 } } I also tried: refinement.pdb_interpretation { apply_cif_link { data_link = DRG-CYS residue_selection_1 = chain A and resname DRG and resid 10 and altloc A residue_selection_2 = chain B and resname CYS and resid 20 } } After refinement (with a corresponding data_link.eff file defining the bonds and angles), the thiol of the other conformer that was facing away from the cysteine moves out of the electron density and now faces the cysteine (still about ~5.7 A away). Is there a way to define a link between only one conformer and have the other refine independently? I did try refinement without the above link definition (relying on the proximity of the two cysteines and automatic linking in phenix) but some of the bond and dihedral angles showed higher deviation from ideal values expected for disulfide bonds. Thanks, Chris