Date: Mon, 24 Jun 2002 10:24:30 -0700 (PDT)
From: Thomas Cheatham
Subject: Re: ptraj problem



> I have a script for calculating the atomic fluctuations:
...
> trajin ../../../${fname}_md00.mdcrd.Z
> trajin ../../../${fname}_md01.mdcrd.Z
> center :1-142 mass origin
> image origin :1-142
>
> atomicfluct out fluct_byatom.dat :126-142
> atomicfluct out fluct_byres.dat :126-142 byres
...
> results for the atomic flucuations. The question is why?

The atomicfluct command is general in that it does not assume a fit of the
coordinates or a reference frame at zero or in other words, it is just a
raw calculation of atomic positional fluctuations. To get the behavior
you would expect, you need to put the molecule of interest into a common
reference frame, such as after RMS fitting it.

trajin ../../../${fname}_md00.mdcrd.Z
trajin ../../../${fname}_md01.mdcrd.Z
center :1-142 mass origin
image origin center

rms first mass out rms_126-142.dat :126-142
atomicfluct out fluct_byatom.dat :126-142
atomicfluct out fluct_byres.dat :126-142 byres

If you want to then get the RMS data for individual residues, do this
after the atomicfluct command.

*** Also, note that there was a bug in the Bfactor option of atomicfluct
that has been recently fixed. The Bfactor is not simply the atomic
fluctuations multipied by a constant but the sqrt() of the fluctuations
multiplied by a constant. A new version of ptraj which is updated with
this fix (and also with a new option to randomly swap ion and solvent
positions and more support for AMBER7 prmtop's) is available at

http://www.chpc.utah.edu/~cheatham/software.html

--tom