[gmx-users] question about energy minimization

[Justin A. Lemkul]

On 7/5/12 10:52 AM, reisingere at rostlab.informatik.tu-muenchen.de wrote:
> Hi everybody,
> I want to minimize the energy of my structure.
> I use the following mdp file:
>
> define                  = -DPOSRES
> integrator              = steep
> emtol           = 10
> nsteps          = 5000
> nstenergy               = 1
> energygrps              = System
> coulombtype             = PME
> rcoulomb                = 0.9
> rvdw                    = 0.9
> rlist           = 0.9
> fourierspacing  = 0.12
> pme_order               = 4
> ewald_rtol              = 1e-5
> pbc                     = xyz
>
> But it nevers runs till the end. It allways stops before the end and I get
> no pdb file. When I use nsteps = 2000 there is no problem and I receive a
> pdb file but it says that there were not enough steps to get under a
> specific force value.
>
> My question is: is there a possibility to make the minimization run
> longer? Can I set a parameter that makes it run longer?
>

Set nsteps to either a very high value or to -1 (which says run infinitely 
long).  Note that several factors are working against you here, including the 
use of position restraints.  These disfavor the motion of whatever the 
restrained atoms are, which can impede energy minimization.  Also, assuming 
you're using single-precision Gromacs, an emtol of 10 is very low and unlikely 
to be achievable, especially using a single run of steepest descent EM.

-Justin

-- 
========================================

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================