[gmx-users] Energy Minimisation and Equilibration Problems

[Bruce D. Ray]

On Thursday, August 6, 2009 at 11:58:20 AM, Nancy <nancy5villa at gmail.com> wrote:

>  The energies simply do not seem to come down any further within several thousand steps.
> 
> I
start with a .mol2 file which contains the structure of ethylene glycol
(ethanediol).
> These are the commands that I use to set up and run the
minimisation:
> 
> $ .../topolbuild1_2_1/src/topolbuild -n ethanediol -dir .../topolbuild1_2_1/dat/gromacs -ff gmx53a6
> 
> the above command outputs the following files:
> 
> ethanediol.gro
> ethanediol.log
> ethanediolMOL.mol2
> ethanediol.top
> ffethanediol.itp
> posreethanediol.itp

Those are the correct files to be generated for the topolbuild command given.
Are the parameters blank in any of the lines in ethanediol.top?  Are there
any lines in ethanediol.log that do not involve nonpolar hydrogens that
have asterisks instead of parameter values?

> I then proceed to enlarge the box and solvate the molecule:
> 
> $ editconf -f ethanediol.gro -o ethanediol_box.gro -box 5 5 5
> 
> $ genbox -cp ethanediol_box.gro -cs spc216.gro -o ethanediol_box.gro -p ethanediol.top -shell 1

Is this a shell of solvent around the molecule rather than a box of solvent containing
the molecule of interest?  How is that to work with vacuum outside the shell?

> I then use grompp to configure the minimisation:
> 
> $ grompp -f minim.mdp -c ethanediol_box.gro -p ethanediol.top -o em.tpr
> 
> This is my .mdp file for minimisation:
> 
> ===============================
> define        = -DPOSRE

Why position restraints when doing energy minimization?

> 
integrator    = steep    
> emtol        = 10.0  
> emstep          = 0.01  
> nsteps        = 2000     

Why so few steps when the emtol is this small.  For this emtol, I would expect
to need 10000 to 20000 steps.

> nstlist        = 1    
> ns_type        = grid
> rlist        = 1.0    
> coulombtype    = PME    
> rcoulomb    = 1.0
> 
rvdw        = 1.3    
> pbc        = xyz 
> 
> pme_order    = 4
> constraints     = none
> 
> 
> nstxout        = 1
> nstvout        = 1
> nstenergy    = 1
> nstlog        = 1
> nstcomm         = 1
> Tcoupl          = no
> 
Pcoupl          = no
> gen_vel         = no
> ===============================

I've been told that the last three lines are not needed for energy minimization.

> $ mdrun -v -deffnm em
> 
> the minimisation runs without error and these are the last few lines of output:
> 
> ===========================================
> 
Step= 1990, Dmax= 1.1e-05 nm, Epot= -1.76427e+04 Fmax= 1.25524e+04, atom= 304
> Step= 1991, Dmax= 1.3e-05 nm, Epot= -1.76433e+04 Fmax= 3.13881e+04, atom= 484
> Step= 1992, Dmax= 1.6e-05 nm, Epot= -1.76443e+04 Fmax= 2.11785e+04, atom= 484
> 
Step= 1993, Dmax= 1.9e-05 nm, Epot= -1.76445e+04 Fmax= 4.22489e+04, atom= 484
> Step= 1994, Dmax= 2.3e-05 nm, Epot= -1.76456e+04 Fmax= 3.33021e+04, atom= 484
> Step= 1996, Dmax= 1.4e-05 nm, Epot= -1.76468e+04 Fmax= 1.26110e+04, atom= 304
> 
Step= 1997, Dmax= 1.6e-05 nm, Epot= -1.76474e+04 Fmax= 3.81619e+04, atom= 484
> Step= 1998, Dmax= 2.0e-05 nm, Epot= -1.76486e+04 Fmax= 2.68240e+04, atom= 484
> Step= 2000, Dmax= 1.2e-05 nm, Epot= -1.76496e+04 Fmax= 1.26501e+04, atom= 304
> 
> writing lowest energy coordinates.
> 
> Steepest Descents did not converge to Fmax < 10 in 2001 steps.
> Potential Energy  = -1.7649566e+04
> Maximum force     =  1.2650057e+04 on atom 30
> Norm of force     =  1.7544460e+03
> 
===========================================
> 
> If
I proceed to equilibration after doing the above, the water molecules
simply fly apart
> (although not immediately).  Additionally, it seems
that there are no interactions between
> the waters during equilibration.

Really no point in proceeding further until the energy minimization issues are solved.

> I have tried to run the minimisation for a larger number of steps,
but it does not help.
> I have also tried to delete individual water
molecules from the structure files, but doing
> so simply causes the
minimisation to "fixate" on another two molecules.  I am not sure
> what
values of the energies are reasonable for this system, and how to
minimise it further. 
> Please advise.

I've never used the -shell with genbox and don't know what effect that would have on
energy minimization.

-- 
Bruce D. Ray, Ph.D.
Associate Scientist, and Operations Director
NMR Center
IUPUI
Physics Dept.
402 N. Blackford St.
Indianapolis, IN  46202-3273


      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20090806/99c8be72/attachment.html>