[cctbxbb] Best way to compare two structures using cctbx?

[Pavel Afonine]

Hi Jan,

there is no dedicated tool that would do it in one go, but I think you 
can put something together from pieces scattered across cctbx and phenix 
code. Oh and by the way I don't know what is "direct space atomic 
distances "...

First I would superpose two structures using a call like this (copied 
from phenix/phenix/command_line):

from scitbx.math import superpose
superpose.least_squares_fit(reference_sites = fixed_sites, other_sites = 
moving_sites)
rmsd = fixed_sites.rms_difference(moving_sites)

this assumes structures are aligned.

Then just loop over bonds and collect bond distances into flex double 
array and then compute call .min_max_mean()...

Pavel


On 3/4/13 4:06 AM, Jan Marten Simons wrote:
> Hi,
>
> I think I've seen some code in cctbx which tries to put the similarity of two
> crystal structures into a number. Am I right on this?
>
> I want to have something similar:
>
> I'm interested in comparing the direct space atomic distances of two
> structures sharing the same unit cell parameters and space group and
> scatterers but different atomic coordinates for those. Now I'd like to have a
> mapping where I could compare the bond lengths between all scatterers which
> should eliminate any movements which might also be due to a shift in the
> origin of a P 1 cell. Also I'd like to have a value for mean, minimal and
> maximal displacements.
> The purpose of this is to check how good a structure approximation matches a
> reference structure.
>
> How would you suggest to take on this task? Or is there even some code related
> to this?
>
> Thanks in advance,
>
> Jan
> _______________________________________________
> cctbxbb mailing list
> cctbxbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/cctbxbb