Hi all,

We have been finalising a structure with a ligand complexed to heme. I have been checking the close to final structure and I noticed that the ligand looks distorted away from what one would expect.  We use Phenix.refine

We initially used the grade server for our restraints files. I also just used elbow to generate a new one. Looking at both, the restraints look reasonable (similar in both) however the initial output pdb from elbow is distorted further. The angles /lengths looked reasonable and are similar to small molecule structures (CCDC).

The planarity of a bi-phenyl moiety is very distorted.

Are we missing something?

Do we need to tell phenix.refine to not over refine the model?

Should elbow give a distorted model?

Apologies if this is all online but my uni cant seem to access phenix.org!

Thanks in advance

Joel

Joel Tyndall | BSc(Hons) PhD

Associate Professor in Medicinal Chemistry
School of Pharmacy | He Rau Kawakawa
University of Otago | Te Whare Wānanga o Otāgo

PO Box 56 9054

Dunedin | Ōtepoti

New Zealand | Aotearoa

Ph: 64 3 479 7293
Skype: jtyndall

Website | pharmacy.otago.ac.nz