[phenixbb] Heme geometry restraint

Paul Emsley pemsley at mrc-lmb.cam.ac.uk
Tue Jul 21 11:13:49 PDT 2020

On 20/07/2020 01:20, Yangqi Gu wrote:
> Dear all,
> I have a question regarding the heme c geometry restraint. I am aware 
> that heme c could undergo distortion when bound to protein, but I am 
> not sure at what resolution I could claim such distortion (such as 
> ruffling/bending/doming). I have a map at 3.6 A resolution and 
> whenever I did the refinement in phenix with real space refinement, 
> some of the hemes are distorted and I am wondering if it is justified 
> to leave it that way or do I have to impose the flat plane geometry. 
> If so, how should I proceed to impose such geometry? Thank you in advance!

During my PhD I was fortunate enough to sit down at the graphics (a E&S 
PS390) with Michael Rossmann to show him my current Hb model. He 
expressed some scepticism about the way I had modelled the heme. "Make 
it flat and show me the difference map" he said. The point being that if 
it truly was  buckled then we'd see several significant difference map 
peaks (at say 3.0 rmsd) around the pyrroles (not the Iron atom) 
consistent with the buckled model.


More information about the phenixbb mailing list