Hello,
I'll give a couple of comments where I can below:
On Sun, May 27, 2012 at 11:31 PM, aa aa <gaussian22.gmail.com> wrote:
> Dear Sir/Madam,
>
> I want to do the RMSF analysis for my system (with ~ 1000 protein
> residues). I did the simulation for 50 ns. I checked that the RMSD and
> gyrate radius are converted after 30ns. So, I told the 30 to 50 ns mdcrd
> files for the RMSF analysis.
>
Bear in mind that these are just single numbers. RMSD can be especially
deceptive. 2 structures with an RMSD of 5 Å from one structure can
themselves be completely different (it's not at all unusual for them to
have an RMSD of 5 Å or so from each other as well). Even testing that the
radius of gyration converges does not say enough about converging to a
particular structure -- especially for certain values of the radius of
gyration.
I found the RMSF obtained from 30-50, 40-50 and 30-40 are different. The
> max peak appears at same domain, but the relative intensity of the peaks at
> different domains are different.
>
> Would you mind to teach me whether I should still conclude the RMSF
> obtained from 30-50 are reliable? Can I still claim that the simulation is
> stable after 30?
>
You don't give many details, so it's hard to really answer your questions.
That said, it is not at all unusual for 2 different chunks of 10 ns to
explore very different parts of phase space (even if they start from the
same point with different velocities!) Also, what do you mean by
"reliable"? Since RMSF is a simple calculation, I would say you can rely
on the values ptraj or cpptraj give you, and unless you made a mistake with
your simulation (which only you would know).
Also, given the complexity of biomolecular systems, this behavior could be
perfectly normal for some systems and completely abnormal for others. If
you are exploring large scale molecular motions that take significantly
longer than 10 ns to explore, then your results seem perfectly reasonable
to me. The analysis you should do depends on exactly what questions you
are asking, as well (which you didn't share with us, either).
HTH,
Jason
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue May 29 2012 - 10:00:03 PDT
