Re: [AMBER] Current rmsd from reference in targeted MD

From: Carlos Simmerling <carlos.simmerling.gmail.com>
Date: Wed, 4 Oct 2017 09:04:55 -0400

you can't generate forces for the first step of MD without doing the
overlap calculation, so it will be done regardless of ntwx or ntpr.

On Wed, Oct 4, 2017 at 8:45 AM, Sonia Ziada <sonia.ziada.gmail.com> wrote:

> Or it can be that the change is so small after 1 step simulation that the
> rmsd is still equal to 0.
> Because if the first rmsd calculation is done on the inpcrd, the first
> value should be equal to zero even for ntwx=2 and ntpr=2 simulation. Or am
> I wrong ?
>
> *ZIADA Sonia*
> PhD Student, Structural Bioinformatics & Chemoinformatics
> Institut de Chimie Organique et Analytique (ICOA), UMR CNRS-Université
> d'Orléans 7311
> Rue de Chartres, 45067 Orléans, France
> T. +33 238 419 939
>
> 2017-10-04 14:38 GMT+02:00 Carlos Simmerling <carlos.simmerling.gmail.com
> >:
>
> > as you can tell from the rmsd value of 0, the first rmsd calculation is
> > done on your inpcrd.
> >
> > On Wed, Oct 4, 2017 at 8:25 AM, Sonia Ziada <sonia.ziada.gmail.com>
> wrote:
> >
> > > Ok I did it and the result is surprising :
> > >
> > > ### 1) From the .mdout file:
> > > Current RMSD from reference: 0.000
> > > Current RMSD from reference: 0.017
> > > Current RMSD from reference: 0.034
> > > Current RMSD from reference: 0.049
> > > Current RMSD from reference: 0.063
> > > Current RMSD from reference: 0.075
> > > Current RMSD from reference: 0.085
> > > Current RMSD from reference: 0.094
> > > Current RMSD from reference: 0.103
> > > Current RMSD from reference: 0.110
> > >
> > > ### 2) From cpptraj (rmsd.dat file)
> > > #Frame RMSD_00002
> > > 1 0.0172
> > > 2 0.0337
> > > 3 0.0491
> > > 4 0.0630
> > > 5 0.0752
> > > 6 0.0853
> > > 7 0.0942
> > > 8 0.1027
> > > 9 0.1106
> > > 10 0.1178
> > >
> > > Does it mean that the first frame in the trajectory does not correspond
> > to
> > > the first step of the simulation but rather the second OR that the
> first
> > > block in the mdout (NSTEP=1) does not correspond to the first step of
> the
> > > simulation but rather to the step 0, that is the restart ?
> > >
> > > Thank you.
> > >
> > > *ZIADA Sonia*
> > > PhD Student, Structural Bioinformatics & Chemoinformatics
> > > Institut de Chimie Organique et Analytique (ICOA), UMR CNRS-Université
> > > d'Orléans 7311
> > > Rue de Chartres, 45067 Orléans, France
> > > T. +33 238 419 939
> > >
> > > 2017-10-04 12:30 GMT+02:00 Carlos Simmerling <
> > carlos.simmerling.gmail.com
> > > >:
> > >
> > > > Can you try writing structures every step?
> > > >
> > > > On Oct 3, 2017 5:05 PM, "Sonia Ziada" <sonia.ziada.gmail.com> wrote:
> > > >
> > > > > Dear Carlos,
> > > > >
> > > > > Thank you for your reply.
> > > > >
> > > > > - I am writing the trajectory (netcdf file) and the mdout every 2
> > steps
> > > > > (ntwx=2 and ntpr=2), so the structures used in cpptraj should
> > > correspond
> > > > > exactly to those in the MD run.
> > > > > - I am sure that the reference used for the fitting in the MD run
> is
> > > the
> > > > > same that those I use in cpptraj
> > > > > - I have also checked that the masks are correct (by looking at:
> the
> > > > header
> > > > > of the mdout, the output of cpptraj and by using ambmask program)
> > > > > - Finally, I have a looked at the ene.F90 sander script containing
> > the
> > > > > subroutine fitrms (that calcutes the current rmsd after the fitting
> > > step)
> > > > > to ensure that the rmsd is mass weighted.
> > > > >
> > > > > I do not understand the difference in rmsd values between cpptraj
> and
> > > the
> > > > > mdout file. By calcutating the rmsd with cpptraj, I assumed that
> the
> > > > > alignment algorithm is the same in cpptraj as in sander. But is
> this
> > > > true?
> > > > >
> > > > >
> > > > > *ZIADA Sonia*
> > > > > PhD Student, Structural Bioinformatics & Chemoinformatics
> > > > > Institut de Chimie Organique et Analytique (ICOA), UMR
> > CNRS-Université
> > > > > d'Orléans 7311
> > > > > Rue de Chartres, 45067 Orléans, France
> > > > > T. +33 238 419 939
> > > > >
> > > > > 2017-10-03 18:44 GMT+02:00 Carlos Simmerling <
> > > > carlos.simmerling.gmail.com
> > > > > >:
> > > > >
> > > > > > are you certain that the structures that you are using in ptraj
> > > > > correspond
> > > > > > exactly to those in the MD run? for example, are you writing
> every
> > > > step?
> > > > > > and are you using netcdf format instead of the low precision
> > > trajectory
> > > > > > files?
> > > > > >
> > > > > > On Tue, Oct 3, 2017 at 7:36 AM, Sonia Ziada <
> sonia.ziada.gmail.com
> > >
> > > > > wrote:
> > > > > >
> > > > > > > Dear Amber users,
> > > > > > >
> > > > > > > I am having trouble understanding how the RMSD is calculated in
> > > > > targeted
> > > > > > MD
> > > > > > > simulations.
> > > > > > >
> > > > > > > I am using tgtfitmask (:Mask1) and tgtrmsmask (:Mask2).
> > > > > > > I try to reproduce the sander output's targeted rmsd results
> > (those
> > > > > > > available in the mdout file: “current rmsd from reference”)
> using
> > > > > cpptraj
> > > > > > > with the following commands:
> > > > > > >
> > > > > > > ### cpptraj script ---------------------------------
> > > > > > > parm topo.top
> > > > > > > trajin traj.nc
> > > > > > > reference reference.rst
> > > > > > > rms reference :Mask1
> > > > > > > rms reference mass :Mask2 out rmsd.dat nofit
> > > > > > > run
> > > > > > > quit
> > > > > > > --------------------------------------------------------
> > > > > > >
> > > > > > >
> > > > > > > *But I get different results: *
> > > > > > > ### 1) From the .mdout file:
> > > > > > > Current RMSD from reference: 0.017
> > > > > > > Current RMSD from reference: 0.049
> > > > > > > Current RMSD from reference: 0.075
> > > > > > > Current RMSD from reference: 0.094
> > > > > > > Current RMSD from reference: 0.110
> > > > > > >
> > > > > > > ### 2) From cpptraj (rmsd.dat file)
> > > > > > > 1 0.0337
> > > > > > > 2 0.0630
> > > > > > > 3 0.0853
> > > > > > > 4 0.1027
> > > > > > > 5 0.1178
> > > > > > >
> > > > > > > As far as I can conclude with these results, the TMD's
> alignment
> > > > sceme
> > > > > > > seems different from cpptraj's alignment sceme. How can I
> > reproduce
> > > > the
> > > > > > > rmsd of the mdout using cpptraj ?
> > > > > > > Thank you.
> > > > > > >
> > > > > > >
> > > > > > > *ZIADA Sonia*
> > > > > > > PhD Student, Structural Bioinformatics & Chemoinformatics
> > > > > > > Institut de Chimie Organique et Analytique (ICOA), UMR
> > > > CNRS-Université
> > > > > > > d'Orléans 7311
> > > > > > > Rue de Chartres, 45067 Orléans, France
> > > > > > > T. +33 238 419 939
> > > > > > > _______________________________________________
> > > > > > > AMBER mailing list
> > > > > > > AMBER.ambermd.org
> > > > > > > http://lists.ambermd.org/mailman/listinfo/amber
> > > > > > >
> > > > > > _______________________________________________
> > > > > > AMBER mailing list
> > > > > > AMBER.ambermd.org
> > > > > > http://lists.ambermd.org/mailman/listinfo/amber
> > > > > >
> > > > > _______________________________________________
> > > > > AMBER mailing list
> > > > > AMBER.ambermd.org
> > > > > http://lists.ambermd.org/mailman/listinfo/amber
> > > > >
> > > > _______________________________________________
> > > > AMBER mailing list
> > > > AMBER.ambermd.org
> > > > http://lists.ambermd.org/mailman/listinfo/amber
> > > >
> > > _______________________________________________
> > > AMBER mailing list
> > > AMBER.ambermd.org
> > > http://lists.ambermd.org/mailman/listinfo/amber
> > >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Oct 04 2017 - 06:30:03 PDT
Custom Search