[gmx-users] fitting of the RMSD matrix

[Mike Nemec]

Dear GROMACS mailing list,

I want to calculate a 2D RMSD matrix using g_rms of gromacs v4.6.7

(1) between all frames of one trajectory and

(2) between all frames of two different trajectories.

I figured out, that this is possible by

     g_rms -f traj.xtc -s traj.tpr -m RMSD_matrix.xpm -bin RMSD_matrix.dat

either specifying a second trajectory by -f2 or not, where the RMSD 
matrix is obtained in binary format in the installed precision 
(RMSD_matrix.dat). Unfortunately, the manual gives only the information 
about the RMSD traces, that they are fitted to the reference, but I did 
not found any information about the RMSD matrix generation.

My question concerns the fitting/superimposing:

Is the RMSD matrix calculated by

(1) fitting all structures to the reference submitted by -s

(2) or by a pair-wise fit between all structures for which then the RMSD 
values are calculated?

If it is not done by a pair-wise fit, is there an option to achieve a 
pair-wise fitted RMSD matrix?

Thank you in advance,

     Mike