# [gmx-users] energy minimization - double precision

Fri May 18 04:57:15 CEST 2007

```Dear users,
I'm trying to minimize the energy of a protein in vacuum. I have restrained
the backbone with posre.itp generated using genpr. when i try to minimize
the energy of this protein using single precision grompp followed by an
mdrun, it does not converge even after 600 steps and returns saying that it
could not get to the requested minimum energy, and suggested that i use
double precision for the same. so, i compiled gromacs with double precision
and tried the same and i get the following. Could you help me solve this?
Sometime back, venky wrote to the list that he would delete some atoms that
clash. Considering i would not want to do that, could you suggest an
alternative?

the error i get is,
********************************************************************************************
Back Off! I just backed up wt_em.edr to ./#wt_em.edr.3#
Low-Memory BFGS Minimizer:
Tolerance (Fmax)   =  1.00000e+01
Number of steps    =       500000
Using 20 BFGS correction steps.

F-max             =          inf on atom 1
F-Norm            =          inf

Back Off! I just backed up wt_em.trr to ./#wt_em.trr.3#
Step 0, Epot=-4.939505e+04, Fnorm=      inf, Fmax=      inf (atom 1)
Step 1, Epot=         nan, Fnorm=      nan, Fmax=0.000e+00 (atom 0)

writing lowest energy coordinates.

Back Off! I just backed up wt_em.gro to ./#wt_em.gro.3#

Low-Memory BFGS Minimizer converged to Fmax < 10 in 1 steps
Potential Energy  =                   nan
Maximum force     =  0.00000000000000e+00 on atom 0
Norm of force     =                   nan

gcq#2: "Jesus Built My Hotrod" (Ministry)

*******************************************************************************************
and this is my mdp file
*******************************************************************************************
cpp                      = /lib/cpp
define                   = -DPOSRES -DFLEX_SPC

; RUN CONTROL PARAMETERS
integrator               = l-bfgs
tinit                    = 0
dt                       = 0.001
nsteps                   = 500000
init_step                = 0

comm-mode                = Angular
nstcomm                  = 1
comm-grps                =

; LANGEVIN DYNAMICS OPTIONS
bd-fric                  = 0
ld-seed                  = 1993

; ENERGY MINIMIZATION OPTIONS
emtol                    = 10
emstep                   = 0.001
nbfgscorr                = 20

; OUTPUT CONTROL OPTIONS
nstxout                  = 1000
nstvout                  = 1000
nstfout                  = 0
nstcheckpoint            = 1000
nstlog                   = 1000
nstenergy                = 1000

nstxtcout                = 0
xtc-precision            = 100
xtc-grps                 = Protein
energygrps               = Protein

; NEIGHBORSEARCHING PARAMETERS
nstlist                  = 1
ns-type                  = Simple
pbc                      = no

rlist                    = 1
domain-decomposition     = no

; OPTIONS FOR ELECTROSTATICS AND VDW
coulombtype              = Cut-off
rcoulomb         = 1
epsilon-r                = 1

vdw-type                 = Cut-off
rvdw-switch              = 0
rvdw                     = 1
DispCorr                 = No

; OPTIONS FOR BONDS
constraints              = none
constraint-algorithm     = Lincs
unconstrained-start      = no
lincs-order              = 4
lincs-iter               = 2
lincs-warnangle          = 30
morse                    = no
*****************************************************************************************

Kota.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20070518/b4132a66/attachment.html>
```