[gmx-users] rmsd between different monomers

[Mark Abraham]

On 9/05/2010 10:01 PM, Anupam Nath Jha wrote:
>
> Ok.
>
> But when I run this command
> g_rms -s setp_0.pdb -f setp_0.pdb -n ca_ind.ndx -o out
>
> it ask me groups... and that i thought is for fitting///
>
> for example -
>
> ------------------------
> Select group for least squares fit
> Group     0 (C-alpha_chain1) has   249 elements
> Group     1 (C-alpha_chain2) has   249 elements
> Group     2 (C-alpha_chain3) has   249 elements
> Group     3 (C-alpha_chain4) has   249 elements
> Select a group: 0
> Selected 0: 'C-alpha_chain1'

So group 0 in *each* structure is used for fitting the two structures to 
each other...

> Select group for RMSD calculation
> Group     0 (C-alpha_chain1) has   249 elements
> Group     1 (C-alpha_chain2) has   249 elements
> Group     2 (C-alpha_chain3) has   249 elements
> Group     3 (C-alpha_chain4) has   249 elements
> Select a group:1
> Selected 1: 'C-alpha_chain2'

... and group 1 in *each* structure is then used for the RMSD calculation.

The above is *not* fitting group 0 to group 1, and then measuring the 
RMSD of group 0 w.r.t. group 1, which sounds about what you want to do. 
You need my editconf/trjconv approach for that.

It's up to you to determine whether groups 0 and 1 should intersect. 
This allows you to (say) do an all-C-alpha fit, and then measure only 
the part of interest. In your case, as Justin said, since the two input 
structures are the same, you get zero RMSD regardless what else you do.

Notice how giving the command line, inputs and outputs has led to 
getting useful feedback, rather than the last 10 emails' worth of 
frustration because nobody knows what the other is talking about :-)

Mark