Re: Question about equilibration

From: Gianluca Interlandi (gianluca_at_u.washington.edu)
Date: Wed Jun 15 2011 - 14:03:57 CDT

Why not plotting the RMSD from the X-ray structure for the total complex
and each helix independently?

Gianluca

On Wed, 15 Jun 2011, Richard Wood wrote:

> Right, as one can have a "good" RMSD, but a terrible structure.
>
> Richard
>
> ___________________________________________________________________________________________
> From: johan strumpfer <johanstr_at_ks.uiuc.edu>
> To: Richard Wood <rwoodphd_at_yahoo.com>
> Cc: snoze pa <snoze.pa_at_gmail.com>; namd-l_at_ks.uiuc.edu
> Sent: Wed, June 15, 2011 1:22:56 PM
> Subject: Re: namd-l: Question about equilibration
>
> Indeed, hence my suggestion: in reaching a plateau in the rmsd you
> find a stable structure to start from for 'production' runs / use as
> point of comparison. Of course, the difference between this structure
> and the crystal structure is possibly also of interest - it really
> depends on the investigation.
>
> Cheers,
> Johan
>
> On 6/15/11, Richard Wood <rwoodphd_at_yahoo.com> wrote:
> > The reason I say this is because a crystal structure of a protein (which I
> > believe he's starting from) is not a realistic structure, so if you're going
> > to
> > compare an equiibrated structure from a dynamics run to that, you could be
> > asking for trouble.
> >
> > Richard
> >
> >
> >
> >
> > ________________________________
> > From: johan strumpfer <johanstr_at_ks.uiuc.edu>
> > To: Richard Wood <rwoodphd_at_yahoo.com>
> > Cc: snoze pa <snoze.pa_at_gmail.com>; namd-l_at_ks.uiuc.edu
> > Sent: Wed, June 15, 2011 1:10:10 PM
> > Subject: Re: namd-l: Question about equilibration
> >
> > The total energy stabilises usually before the structure has
> > stabilised. It has happened to me quite often that the total energy
> > fluctuates about a well defined mean by 1 ns, but the rmsd is not yet
> > stable and only plateau's after 5-10 ns. It is very much dependent on
> > your starting set up.
> >
> > Cheers,
> > Johan
> >
> > On 6/15/11, Richard Wood <rwoodphd_at_yahoo.com> wrote:
> >> Wouldn't the total energy be a better deterministic tool as to whether or
> >> not
> >> you have equilibrated?
> >>
> >> Richard
> >>
> >>
> >>
> >>
> >> ________________________________
> >> From: johan strumpfer <johanstr_at_ks.uiuc.edu>
> >> To: snoze pa <snoze.pa_at_gmail.com>
> >> Cc: namd-l_at_ks.uiuc.edu
> >> Sent: Wed, June 15, 2011 11:48:06 AM
> >> Subject: Re: namd-l: Question about equilibration
> >>
> >> Snoze,
> >>
> >> You are the only one who can really decide that. Proteins move
> >> stochastically,
> >> they do not stay in one conformation, even simple helices. They are
> >> influenced
> >> by the other proteins and the environment surrounding them, which is
> >> usually
> >> not
> >> symmetric. As I said before, equilibration runs need to be longer usually:
> >> look
> >> at your rmsd as a function of time and when it reaches a plateau, then you
> >> may
> >> be in a better position to say whether your structure is equilibrated.
> >> Whatever
> >> kinks or deformation are present, you will need to look at the structure,
> >> at
> >> the
> >> interactions and justify them for yourself.
> >>
> >> Cheers,
> >> Johan
> >>
> >>----------------------------------------------------------------------------------------
> --------------
> >>-
> >>
> >> Johan Strumpfer:        johanstr_at_ks.uiuc.edu
> >> www.ks.uiuc.edu/~johanstr
> >> Theoretical and Computational Biophysics Group
> >> 3115 Beckman Institute
> >> University of Illinois at Urbana-Champaign
> >> 405 N. Mathews
> >> Urbana, IL 61801, USA
> >>----------------------------------------------------------------------------------------
> --------------
> >>-
> >>
> >>
> >>
> >>
> >> On Wed, Jun 15, 2011 at 11:38 AM, snoze pa <snoze.pa_at_gmail.com> wrote:
> >>> Hi John and NAMD users,
> >>>
> >>>  Thanks for your reply. My question was related to the deformation of
> >>> protein during equilibration. First part of the equilibration for 1ns,
> >>> when
> >>> I used harmonic constrain went well. I don't see  big structural changes.
> >>> But when I remove the harmonic constraint and equilibrate further the
> >>> system
> >>> for 1ns then till 600 ps i don't see any big change in the structure
> >>> except
> >>> the large kink that deform the top part of the two helices in last 300
> >>> ps.
> >>> All three helices have similar sequences and they are symmetric in
> >>> structure. Theoretically all three will show deformed kinked region in
> >>> top
> >>> but only two are showing it.
> >>>
> >>> So question was: is it okey to have such large movement during
> >>> equilibration(knowing the fact that protein is symmetric)?  or do you
> >>> think
> >>> I have to run it more to see if there are more structural changes before
> >>> running full MD simulation.
> >>>
> >>> Thank you,
> >>>
> >>> S
> >>>
> >>>
> >>> On Wed, Jun 15, 2011 at 11:27 AM, johan strumpfer <johanstr_at_ks.uiuc.edu>
> >>> wrote:
> >>>>
> >>>> It looks like a totally reasonable structure to me, but I have no idea
> >>>> what you are working with, how you've set the system up or what it is
> >>>> expected to do. You are the only one that can really decide whether
> >>>> the deformation makes sense and what it's implications are.
> >>>>
> >>>> It is not possible for someone to look at your structure and say
> >>>> whether some deformation is problematic or not, unless they are very
> >>>> familiar with it or it is clearly a physical impossibility.
> >>>>
> >>>> Cheers,
> >>>> Johan
> >>>>
> >>>>
> >>>>--------------------------------------------------------------------------------------
> ----------------
> >>>>
> >>>>-
> >>>> Johan Strumpfer:        johanstr_at_ks.uiuc.edu
> >>>>  www.ks.uiuc.edu/~johanstr
> >>>> Theoretical and Computational Biophysics Group
> >>>> 3115 Beckman Institute
> >>>> University of Illinois at Urbana-Champaign
> >>>> 405 N. Mathews
> >>>> Urbana, IL 61801, USA
> >>>>
> >>>>--------------------------------------------------------------------------------------
> ----------------
> >>>>
> >>>>-
> >>>>
> >>>>
> >>>>
> >>>> On Wed, Jun 15, 2011 at 11:20 AM, snoze pa <snoze.pa_at_gmail.com> wrote:
> >>>> > Please find the link to check the deformed protein
> >>>> >
> >>>> > http://img10.imageshack.us/i/deform.png/
> >>>> >
> >>>> > On Wed, Jun 15, 2011 at 11:07 AM, snoze pa <snoze.pa_at_gmail.com> wrote:
> >>>> >>
> >>>> >> Dear NAMD Users,
> >>>> >>
> >>>> >>
> >>>> >> I need some explanation regarding my protein equilibration. I see my
> >>>> >> protein gets deformed during the equilibration. Do you think it is
> >>>> >> okey?
> >>>> >>
> >>>> >> To add one more note about the protein deformation. First I used
> >>>> >> harmonic
> >>>> >> constrained during the equilibration. Then the restrain were removed
> >>>> >> and
> >>>> >> protein was again equilibrated. I am getting deformation after
> >>>> >> removing
> >>>> >> harmonic constraints. In this step during first 600 ps it is fine but
> >>>> >> last
> >>>> >> 300 ps the protein is getting deformed.
> >>>> >>
> >>>> >> I will highly appreciate your help.
> >>>> >>
> >>>> >> Thank you
> >>>> >>
> >>>> >> S
> >>>> >>
> >>>> >>
> >>>> >
> >>>> >
> >>>
> >>>
> >
> >
> > --
> > ----------------------------------------------------------------
> > Johan Strumpfer (johanstr_at_ks.uiuc.edu)
> > Theoretical and Computational Biophysics Group
> > 3115 Beckman Institute
> > University of Illinois at Urbana-Champaign
> > 405 N. Mathews
> > Urbana, IL 61801, USA
> >
>
>
> --
> ----------------------------------------------------------------
> Johan Strumpfer (johanstr_at_ks.uiuc.edu)
> Theoretical and Computational Biophysics Group
> 3115 Beckman Institute
> University of Illinois at Urbana-Champaign
> 405 N. Mathews
> Urbana, IL 61801, USA
>
>

-----------------------------------------------------
Gianluca Interlandi, PhD gianluca_at_u.washington.edu
                     +1 (206) 685 4435
                     http://artemide.bioeng.washington.edu/

Postdoc at the Department of Bioengineering
at the University of Washington, Seattle WA U.S.A.
-----------------------------------------------------

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:20:27 CST