[gmx-users] strange RMSD output

Christopher Neale chris.neale at mail.utoronto.ca
Sun Oct 28 20:55:07 CET 2012

Try looking at the RMSD after fitting only the the backbone that is not in this variable loop. 
Since you are fitting to the whole protein, and your entire protein group is larger than your loop, 
your result does not surprise me. I would suggest:

a) Do the fitting based on the backbone but excluding the variable loop
b) Compute the RMSD for the i) whole backbone, ii) backbone excluding the loop, and iii) the backbone of the loop only.

Finally, next time please post all relevant commands that you used. You don't mention using the -fit command in
g_rms and you don;t mention running trjconv to do a preliminary fitting. You make it harder for us to help you 
when you provide incomplete information and we need to guess what you did.


-- original message --


  I've got several individual .pdb file and I use the following command to merge it input .xtc and calculate the rmsd by g_rms:

trjcat -cat -keeplast -f *.pdb -o out.xtc

g_rms -f out.xtc -s 001.pdb -o rmsd.xvg

However, I find a problem with the results. Since most part of each PDB
file overlapped very well only a small loop is different. I first
calculate the rmsd of backbone and got:

@ subtitle "Backbone after lsq fit to Backbone"
     0.0000000    0.0000003
     0.0000000    0.3306260
     0.0000000    0.2545602
     0.0000000    0.3293277
     0.0000000    0.2299789
     0.0000000    0.3216407

Then I calculate the loop region and got:

@ subtitle "Backbone_&_r_120-131 after lsq fit to Backbone"
     0.0000000    0.0000003
     0.0000000    0.1526730
     0.0000000    0.1202507
     0.0000000    0.1449232
     0.0000000    0.1318949
     0.0000000    0.1675032

As we can see, the rmsd of loop region is smaller than the backbone
which I think it should be reversed.......

thank you very much.

More information about the gromacs.org_gmx-users mailing list