[phenixbb] Phenix.refine Restraints: replicating refmac external restaints

Elliot Nelson elliot.nelson at dtc.ox.ac.uk
Thu Jan 19 00:49:37 PST 2017

Hi Pavel,

I'm aiming to restrain residues adjacent to alternate conformers, such that the alternate conformers are not able to significantly drift from initial positions within the refinement cycle.

i.e Residue with two alternate conformers/altlocs A and B




with the bold residues restrained to one another.  Hope this makes sense.


From: Pavel Afonine [pafonine at lbl.gov]
Sent: 18 January 2017 18:28
To: Elliot Nelson; phenixbb at phenix-online.org
Subject: Re: [phenixbb] Phenix.refine Restraints: replicating refmac external restaints

Hi Elliot,

Looking to generate restraints for phenix.refine that are similiar/equivalent to external restraints in REFMAC. From over viewing the documentation i feel the correct way to do this is to generate bond restraints, i.e:

Equivalent of:

exte dist first chain A resi   38 atom  C   alte A second chain A resi   38 atom  C   alte B value 0.0 sigma 0.02 type 1

would be :

I don't know how restraints function in Refmac.

refinment.geometry_restraints.edits {
      bond {
                            action = *add
                            atom_selection_1 = chain A  and resid 38 and altloc A and name C
                            atom_selection_2 = chain A  and resid 38 and altloc B and name C
                            symmetry_operation = None
                            distance_ideal = 0.0
                            sigma = 0.02
                            slack = None

This will define a bond between these two atoms by adding a term to the restraints:

(1/sigma**2)*(distance_model - distance_ideal)**2

Note, you can add as many of such bonds as you with. Also, using similar syntax you can add angles, planes, torsions, etc.

Just out curiosity and if it's not a secret: what are you trying to achieve?


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20170119/8ae60cd9/attachment.htm>

More information about the phenixbb mailing list