[gmx-users] How to get the correct reference frame to run the rmsf?

[Justin Lemkul]

On 6/22/17 12:09 PM, ZHANG Cheng wrote:
> Dear Gromacs,
> I try to use this command to calculate RMSF:
> 
> 
> echo 3 | gmx rmsf -s md_0_1.tpr -f md_0_1_noPBC.xtc -o rmsf_20-30ns.xvg -oq bfac.pdb -res -b 20000 -e 30000
> 
> 
> My simulation lasts for 30 ns, but I only want RMSF for the last 10 ns. It is mandatory to assign a reference frame, so I use "md_0_1.tpr", which gives the frame at time 0 ns.
> 

The structure in -s is only used as a reference for fitting if you supply the 
-fit option, which you have not.  The only mandatory elements read from the .tpr 
are the masses and connectivity/residue information.  RMSF is always computed 
from the average structure in the analyzed time frame.

> 
> However, I think I should use a frame which is a time-averaged position within 20-30 ns. Can I ask how to get this frame?
> 

This is what the -ox option does.  It's not a reference frame, though, and the 
coordinates are going to be unphysical.

> 
> In addition, the command line outputs a PDB file. So what does this conformation represents, time 0, 20 or 30 ns?

Compare the coordinates with those snapshots and see :)

-Justin

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

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

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