Xinghua:

A couple of questions you might ask before reverting to drastic measures:

Can you exclude a covalent bond to the protein?

How confident are you about the fit, orientation and nature of the ligand?

Do you see the same phenomenon in a different dataset, i.e. is it reproducible?

It is better to look for an explanation of your experimental findings than to tweak/restrain the refinement to conform to your expectations.

At this resolution the occupancy will be tightly coupled to B-factors. You could try and refine an overall occupancy for the ligand. What makes you believe that you have partial occupancy? Residual fofc density?

Carsten

From: [email protected] [mailto:[email protected]] On Behalf Of [email protected]
Sent: Monday, August 10, 2015 10:43 PM
To: phenixbb
Subject: [phenixbb] ligand refinement

Hi everyone,

I am refining a structure of protein/ligand complex (2.75 Å) almost to the last stage. However, I found that some atoms of the ligand is too close to the protein (1.2 or 1.3 Å) after refinement.

How should I avoid this? Should I use the "nonbonded_distance_cutoff" parameter? What the value should be set to?

Another question is that at this resolution is it possible for  me to refine the occupancy?

Any suggestion is appreciated.

Best

Xinghua

---

Xinghua Qin Ph.D

Department of Physiology
The Fourth Military Medical University
169# Changlexi Road
Xi'an 710032, China
Tel. 86-29-84772779

E-mail: [email protected]