Dear Mariah and Aaron, As mentioned by others, multiple conformations of inhibitors are quite common. I observe them in 1-5% of the cases. An example is published in J.Med.Chem. (2002), 45:2745, pdb code 1LPK. It may also be that some side chains adopt different positions, depending on the conformation of the inhibitor. I estimate the occupancy from the electron densities of the part of the molecule which has different conformations and here steps of 0.1 are sufficient precise (see Bernard Rupps comment). You could also try to refine group occupancies but then you have to make sure that the sum of the occupancies does not exceed 1.
I do not agree that Refmac is just doing fine. In my hands using Refmac version 5.5.0044 I still observe a weak repulsion between both conformers, which may not be noticable if one only does a few cycles of refinement or one does not look very careful at the results, but which nevertheless compromises the refinement. In this case I still have to resort to CNS. I did not try other refinement programs. Herman -----Original Message----- From: CCP4 bulletin board [mailto:ccp...@jiscmail.ac.uk] On Behalf Of Aaron Oakley Sent: Tuesday, January 13, 2009 1:19 AM To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Ligand binding in multiple conformation >I had a question about flexibility in ligand binding in an enzyme active site. >Is it possible for a substrate/product analogue to bind in more than one conformation in the active site. Yes. It is even possible for portions of a ligand to be disordered and not discernable in electron density. >Since the ligand/enzyme interactions are very specific I am a little confused about this. >Also which program would you use if you have to refine with alternate ligand conformation. REFMAC will do this just fine. Just make sure you have the "A", "B" conformers set in the PDB file. EG: ATOM 72 CA APRO A 7 -2.619 28.983 -0.796 0.62 6.48 C ATOM 73 CA BPRO A 7 -2.226 29.044 -0.847 0.38 5.76 C
