RMSD

hi
please let me know, how can I get the RMSD
of atoms in lammps.

thanks in advance
ramin

|

I was interested in the same calculation, I ended up post-processing.

You might try http://lammps.sandia.gov/threads/msg03088.html.

Please clarify:
"RMSD" means different things to different people.

LAMMPS has an "msd" compute, but I don't think it's what you want.
http://lammps.sandia.gov/doc/compute_msd_molecule.html

I assume you have two different structures with the same number of
atoms, and you want to calculate the root-mean-squared distance
between the positions of the corresponding atoms in the two different
conformations?

....If so, do you want the RMSD computed before or AFTER optimal
supperposition (rotation/translation)? This is the most common
meaning of "RMSD".

Are the two structures you are comparing different snapshots from the
same trajectory. Or are you comparing against a reference structure
(which does not change during the simulation).

In other words: Do you want to compute once for every frame in your
trajectory against a reference structure?
Or do you want to compute an all-against-all comparison between every
pair of frames in your trajectory (clustering or time-correlation-rmsd
analysis).

------------- (pause, to get your hopes up)--------------

....Unfortunately I think the answer is that I don't think LAMMPS or
pizza.py can do any of these things (yet). But a number of other
programs can. These programs are analysis tools are typically
distributed with other simulation programs like gromacs or amber.

I don't use them and I can't remember their names, but hopefully
somebody else can post a follow up message.

You can also use a visualisation program such as VMD, Chimera, pymol,
especially if you only have a single pair of structures to compare.
Unfortunately if you use LAMMPS "dump" format, then these programs
will usually not understand the native LAMMPS "dump" trajectory
format. (However VMD does understand DCD files.) If you save your
files in DUMP format, you have to convert your LAMMPS trajectories to
to PDB or XYZ format to use these tools. I've read that can do this
with pizza.py or VMD/topools (or perhaps lammps2pdb.pl) can convert
lammps dump files into PDB format. (I hate PDB format.)

What I think you are trying to do is quite doable. Hopefully somebody
else can recommend a way to do it easily.
Good luck.

Andrew

PS.: Sometimes "RMSD" is interpreted a different way: Imagine you
have two structures, and for each structure you compute a
(2-dimensional) table of distances between all pairs of atoms in that
structure. Sometimes "RMSD" is interpretted as the RMS-difference
between the numbers in these tables. This usage is much less common
however.

I found this post:
http://lammps.sandia.gov/threads/msg03088.html
The relevant doc page is here:
http://lammps.sandia.gov/doc/compute_displace_atom.html