AMBER Archive (2005)

Subject: Re: AMBER: RMSD Calculations

From: Thomas Cheatham (cheatham_at_chpc.utah.edu)
Date: Mon Dec 05 2005 - 16:50:36 CST


Following up on some of the advice you got today...

> I have a MD simulation of a piece of DNA with a compound in it. I also have a
> MD simulation of the DNA without the compound. I'd like to look at how the
> DNA differs.

If the ligand/compound is after the DNA, then you can use the rms
reference and get the RMSd... The "reference" command in ptraj allows one
to load up a PDB structure (like the average structure) and do a 1-1
comparison of the RMSd. It is OK that the sizes are different (i.e.
ligand omitted) as long as the DNA was first and you are only doing the
comparison on the DNA atoms. ptraj will complain when you use the
reference pdb that the size does not match, but it will work if it is the
same DNA in the same order...

> 1. I am going to get an average structure of the DNA for the last 50 ps
> (where it is stable) and save it as a pdb. (I have a smoothing script for VMD
> to do this.)

trajin dna_md1.mdcrd
trajin dna_md2.mdcrd
strip :WAT
strip @Na+
rms first mass out dna_rms.dat @P,O?P,C?'
average avg_1-2ns.pdb pdb start 1000 stop 2000

> 2. I would like to compare the backbone of the DNA in my trajectory to the
> backbone of the averaged DNA-only pdb.

trajin dnacomplex_md1.mdcrd
trajin dnacomplex_md1.mdcrd
strip :WAT
strip @Na+
reference avg_1-2ns.pdb
rms reference mass out dnacomplex_backbone_rms.dat @P,O?P,C?',O?'
rms reference mass out dnacomplex_base_rms.dat @??

> I'm also worried that it will only calculate the RMSD if there are an
> identical number of atoms in each, which I don't have.

It will work if the atoms to compare are in exactly the same order, i.e.
:1-20 in the reference matches :1-20 in the comparison. This will most
likely be the case if the compound bound is non-covalent and listed AFTER
the DNA in the coordinates...

Good luck.

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu