[gmx-users] Does RMSD only consider the "relative" coordinate changes for the selected group?

[Justin Lemkul]

On 11/23/17 11:20 AM, ZHANG Cheng wrote:
> Dear Gromacs,
> When I calculate the RMSD for the whole protein, I got values mostly from 0.2-0.5 nm. However, when I only calculate for a particular residue (using an index file), the scale is mostly only 0.01-0.02 nm, even for a residue on the loop.
>
>
> My understanding is: when doing the RMSD, the software will align the selected group to the group in the reference, as much as possible. Then the software calculates the root mean square deviation.
>
>
> As a result, though a protein may deviate a lot from its reference structure, if only one residue is selected for RMSD, the relative positions of the atoms within that residue still remain almost the same relative coordination, which makes their RMSD only 0.01-0.02 nm.
>
>
> Can I ask if my understanding is correct?

Yes, but note that fitting a residue to itself is practically useless 
(as you are seeing with your inconsequential residue RMSD values); one 
would more often fit to the whole protein (or backbone, CA, etc) and 
subsequently calculate the RMSD of a given residue if its changes are of 
interest.

>
> I wonder, if I can write some script (e.g. python) to manually extract the coordinates information from various frames in the xtc/trr file?

Sure, but you can also do that with gmx traj -ox.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Assistant Professor
Virginia Tech Department of Biochemistry

303 Engel Hall
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129
http://www.biochem.vt.edu/people/faculty/JustinLemkul.html

==================================================