*** For details on how to be removed from this list visit the *** *** CCP4 home page http://www.ccp4.ac.uk ***
[EMAIL PROTECTED] wrote on 06/01/2006 08:15:06 AM: > *** For details on how to be removed from this list visit the *** > *** CCP4 home page http://www.ccp4.ac.uk *** > > > The resolution is 2.5 A. > > Indeed, there should be no constraint to tie the ligands together, > but why do the 2 compounds overlap exactly (in the identical part) > if I refine only one of them at a time in REFMAC and superimpose the > structures afterwards ?? > Why do they overlap exactly when I use CNX istead? > > Biochemically, I expect them to bind in a similar way, but for sure > there is no guarantee that they will bind exactly the same way (for > the identical part). > > Regards > Christian > Hi Christian - Without seeing pictures or rmsd's, I don't know what you mean by "overlap exactly", but my guess is this: if you only have one molecule to fit against the data, refmac will give you the same result every time for the same model - or in your case, the parts of the models that are the same. However, if you have two superimposed models, refmac now has extra degrees of freedom to try to fit the data more precisely, and so it tweaks the models a bit this way and a bit that way so that their combined structure factors fit the data better than either one could alone. Now, a good deal of this "improved" model is probably fitting noise in your data. But with good data, the noise level should be low enough that the two copies overlap reasonably well. So the question is, how badly are your models failing to overlap? RMSD > 0.1 A? Or are the lines not right on top of each other in O (which could be an rmsd as low as ~0.03 A)? Do the identical parts have significantly different conformations - torsion angles > 20 degrees different, bonds flipped, etc.? Can the different positions of the identical region be explained by the positions of the varying region? I'm not sure why CNX gives tighter overlap than refmac does. Perhaps a harmonic restraint that you didn't realize you'd assigned? At the end of the day, I wouldn't worry too much about the variations if they're small. Just tell your modelers that this is what happens when you work with REAL data! - Matt -- Matthew Franklin phone:(917)606-4116 Senior Scientist, ImClone Systems fax:(212)645-2054 180 Varick Street, 6th floor New York, NY 10014 Confidentiality Note: This e-mail, and any attachment to it, contains privileged and confidential information intended only for the use of the individual(s) or entity named on the e-mail. If the reader of this e-mail is not the intended recipient, or the employee or agent responsible for delivering it to the intended recipient, you are hereby notified that reading it is strictly prohibited. If you have received this e-mail in error, please immediately return it to the sender and delete it from your system. Thank you.
