On Mon, Aug 17, 2015 at 3:53 PM, Krantzman, Kristin D <KrantzmanK.cofc.edu>
wrote:
> Dear All,
> I have a puzzle that I do not understand. I have a system with 5
> molecules that have resid's 1, 2, 3, 4 and 5. When I select one residue and
> have cpptraj calculate the rmsd for a trajectory, the value is very small
> at all time steps, i.e. rms :1 out res1.rms. I get a small value for all
> time steps for all the residues, i.e. rms :2 out res2.rms, etc. However,
> if I calculate the rms for all of the residues (five total) collectively,
> i.e. rms :1-5 resall.rms, the value of rmsd is much larger and increases
> over time. Could anyone explain this discrepancy? I have this when I do
> the calculation on a wrapped trajectory and also when I include the keyword
> noimage. I have searched through the Amber archives and reference manual,
> and I cannot figure out what is incorrect.
>
Nothing is incorrect -- this sounds exactly like what *should* happen.
When you do an "rmsd" command in cpptraj, it automatically applies a
translation and rotation that minimizes the requested RMSD (unless you use
the "nofit" keyword to prevent this).
When you do each molecule independently, it will translate each frame and
rotate them to minimize the RMSD. For a single molecule, this completely
eliminates the effects of translational and rotational diffusion.
However, when you have *multiple* molecules in your RMSD fit, the
translation and rotation eliminated by the fit is for the *combined* system
of all selected molecules. However, their translations and rotations
*relative to each other* cannot be eliminated. As a result, the RMSD
appears much larger in this case. Furthermore, the longer the simulation
runs, the more these molecules diffuse, so the larger the RMSD will
become. What you're observing here is basically the difference between
considering and ignoring diffusion in an RMSD calculation.
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Aug 17 2015 - 13:30:03 PDT
