From: Jérôme Hénin (
Date: Thu Feb 24 2011 - 05:47:36 CST

Hi Alex,

On 23 February 2011 19:51, Alex Liu <> wrote:
> Dear VMD users,
> I have a few questions regarding RMSF feature in VMD:
> 1. I want to calculate RMSF for an atom. I need to align my molecule
> first. I Know that using RMSDTT in VMD can do this for me. So, I just load
> the dcd file and click align (ref is top molecule) in RMSDTT. Then I save as
> coordinated (dcd) and reload in VMD or I can also load the initial structure
> and then load the trajectory file and choose the first structure as ref and
> click align and then save the coordinates of the second structure .Then I
> can use:
>  set outfile [open rmsf.dat w]
>  set sel [atomselect top "name CA"]
>  puts $outfile "[measure rmsf $sel first 1 last 1500 step 1]"
>  close $outfile
> as shared on the website to measure RMSF on that atom. Is this correct?
> Because the RMSF value I am getting is very different from what I expect to
> see.

Note that the first frame is numbered 0 (and the last one is N-1). If
you want all frames, just omit these parameters.

> 2. I Know that using VMD one cannot measure RMSF for a group of atoms
> (example: center of mass of some atoms). Is there any suggestion how to do
> this?

You can choose a 'dummy' atom that you'll move to the center of the group:

set group [...]
set dummy [...]

<loop on frames> {
  $group frame $f
  $dummy frame $f
  $dummy set {x y z} [measure center $group]

Then use measure rmsf on $dummy.