In the best case, you would know something about the enzyme bound equilibrium constant between the two ligand mixtures (see, for example, TM Larsen et al Biochemistry 35, 4349). That would give you a good sense of how to model the occupancies of the active site ligands.
Unfortunately, in many cases, the on-enzyme equilibrium constant is not known and we are left with mixtures of ligands in the active site. I believe that there is no single correct way to then model your structure and (as with many topics) you¹ll get multiple opinions: You can refine the ³predominant² ligand and acknowledge extra difference density that is likely due to other ligands, you can refine a 50/50 mixture and allow the B-values to accommodate some of the differences, or you can refine some other mixture (although I personally am less comfortable with claiming a 35/65 % mixture of your ligands). As you note, the addition of an extra oxygen is not really going to dramatically alter your R-factors so you cannot rely on that to determine the ³correct² model. I think that what you have described is appropriate. It is now important to provide enough information so that your readers can judge any biochemical interpretations that you make. Best wishes, Andy -- Andrew M. Gulick, Ph.D. Hauptman-Woodward Institute On 1/18/08 3:09 PM, "Sun Tang" <[EMAIL PROTECTED]> wrote: > Hello everyone, > > I have a structure of intermediate state in which about half amount of ATP > decomposed to AMP and pyrophosphate. The ATP and AMP + pyrophosphate have > little difference in conformation, sharing the same electron density. > > I just gave them different residue ID and did the TLS and restrained > refinement in CCP4i. It is hard to tell from the R-factor because they are > only a very small part of the whole structure. Can anyone tell whether it is > the correct way to do? > > Any suggestions are greatly appreciated. > > Thank you very much! > > Sincerely, > > Sun Tang > >
