I have never seen a structure with guanine where the map was good enough to see where the hydrogen atoms are so I can't comment directly on your question. I do know that a long time ago Pauling and Corey wrote that the peptide bond was planar with an omega angle of +- 10 deg or so. Over time the fact that the bond was planar was remembered but the error bars seem to have been mostly forgotten. Then when protein diffraction data sets going to high angles of scattering started to become available it was "discovered" that peptide bonds were not as flat as expected and we had to adjust our restraints. My recollection of the last time I did an omega survey in models of sufficient quality to be able to overcome bad restraints the rmsd was about 7 deg. This means it does not take long at all to find a peptide bond with 15 deg deviation from perfect planarity, just as Pauling said. The paper that Pavel recommended shows the same variability for the side chain of Arg which has very similar bonding structure to the peptide bond. Its variability shouldn't have been a surprise, based on the prior experience with peptide bonds, but refinement software was still pushing hard for flat Arg side chains. As a guess, I would say that the planarity imposed by a partial double bond shouldn't be held tighter than the 10 deg rmsd seen in these examples. Those examples, of course, where all non-hydrogen atoms and you are asking about an NH2 group. Hydrogen atoms shouldn't really be making pi bonds so I would think they are even more variable, but as I said I have no experimental data to back that guess. Dale Tronrud P.S. You mention that you are using "cdl". If by that you mean the Conformation Dependent Library, I'll point out that a "cdl" has been created only for protein main chain atoms. I'm afraid that the grant was not renewed for that project so Andy Karplus and I were unable to expand that concept to wider application. To my knowledge there is no CDL for guanine. On 4/27/2026 8:14 AM, [email protected] wrote:
Dear all,
I am trying to refine a high-resolution cryo-EM structure. However, some of the guanine NH2 groups end up out of plane (I used cdl). Is that an expected behavior ?
It seems that may be planarity restraints associated with hydrogens are not strong enough. Is there an option to augment them ? May be also the riding option only involves distance and angle restraints ? This is not clear from the manual to me.
Also, for OH groups like ribose 2'-OH groups, I assume that the riding option only involves distance restraints and that these H-atoms are otherwise free to rotate
Can you please help me out here ?
Thanks, Pascal _______________________________________________ phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s