On Sun, Apr 01, 2018, Mark Rance wrote:
>
> I was wondering if there is a way in cpptraj to align frames such that two
> atoms are superimposed in all frames? I wanted to calculate the
> autocorrelation function of the rotation of a methyl group, so I used the
> rms command to align the frames, specifying in the mask only the two atoms
> that define the symmetry axis of the methyl group. But the output of the
> autocorrelation calculation just had nan in the second column. It seems
> that the rms command wants at least 3 atoms in the mask, as far as I could
> tell.
You are correct: three atoms are needed to make a sensible
superposition. If you wanted to look at methyl rotation in alanine (for
example), you might choose the N,CA and CB atoms, or (perhaps better),
the N,CA,HA,CB and C atoms of the alanine. The former should give you
an exact alignment of CA and CB, but the latter seems more like what you
might want: CA--CB aligned in the local from of the alanine residue
backbone.
Aside from the mathematical problem, if you just specified CA,CB, the
superposition program would not know the difference between rotation of
the hydrogens on CB versus rotation of the entire protein about the
CA-CB bond. (Not very well worded, but I hope you get the point).
I had thought that cpptraj would provide an error message if a user
tried to do an RMS superposition with only two atoms specified...maybe
Dan has more insight here.
....good luck....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Apr 01 2018 - 08:30:02 PDT
