[gmx-users] g_RMS: Mismatch of atom number between topology and reference structure

[Mark Abraham]

On 7/03/2011 7:33 PM, Sergio Manzetti wrote:
> Hello, I am doing a refolding experiment, but I re-experience the same 
> error. First of all, I use pdb2gmx for an input PDB structure, choose 
> OPLS/AA force field and vacuum settings (no waters).
>
> Then I amke a box, and start simulations.
>
> What I want to do is to compare the starting unfolded structure of the 
> simulation and its changes during the sim (all represented as a 
> traj.xtc) to a native folded conformation of the structure. Assuming 
> it to be possible, I thought of using g_rms;
>
> g_rms -f native-folded.gro -s topol.tpr
>
>
> Then I get the same error message:
>
> WARNING: topology has 497 atoms, whereas trajectory has 491
>
> -------------------------------------------------------
> Program g_rms, VERSION 4.5.1
> Source code file: mshift.c, line: 102
>
> Fatal error:
> Molecule in topology has atom numbers below and above natoms (491).
>
> Thing is that both the topology and the trajectory are treated equally 
> at the pdb2gmx step. So how can it be?

You've almost certainly mismatched these files. Only your records can 
make sense of what has gone wrong. The source of this folded form has to 
be treated in an equivalent fashion (e.g. with another invocation of 
pdb2gmx) so that the atom ordering and numbering matches.

Mark