[gmx-users] Re: energy-mimisation-problem

[Justin Lemkul]

On 11/26/12 6:01 AM, SANTU BISWAS wrote:
> On Sun, Nov 25, 2012 at 9:08 PM, SANTU BISWAS <santu.biswas85 at gmail.com> wrote:
>> dear users,
>>
>>                 When performing a energy minimization of a
>> polypeptide(formed by lysine-5-residues) in vacuum box by using
>> Steepest Descent and also Conjugate Gradient methods in gromacs double
>> precision,i noted that GROMACS never converges to emtol values under
>> about 0.00001 kj/mol/nm.I have used the .mdp file which is given below
>>
>>
>>
>> title                    =
>> cpp                      = /lib/cpp
>> ;include                  =-I../top/
>> define                   = -DFLEXIBLE
>>
>> ; RUN CONTROL PARAMETERS =
>> integrator               = steep
>> ; start time and timestep in ps =
>> tinit                    = 0
>> dt                       = 0.001
>> nsteps                   = 1000
>> rlist                    = 0.9
>> rcoulomb                 = 0.9
>> ; Method for doing Van der Waals
>> vdw-type                 = Cut-off
>> ; cut-off lengths
>> rvdw_switch              = 0
>> rvdw                     = 0.9
>> ; Neighbour searching
>> nstlist                                  = 1
>> ; ENERGY MINIMIZATION OPTIONS =
>> emtol                    = 0.00001
>> emstep                   = 0.1
>> nstcgsteep               = 1000
>>

Are you using single or double precision?  Single precision steepest descents is 
unlikely to ever reach such a low emtol.  Double precision and more thorough 
methods (CG and/or L-BFGS) may achieve it, perhaps after several rounds of each.

-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

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