[gmx-users] problem with energy minimization

[hugo verli]

> On Wed, 12 Dec 2001, Geraudis Mustelier wrote:
> 
> >Hi
> >
> >I would like to refine a model obtained with modeling by homology
> technique
> >from X-ray structures. So, I am treating to do an energy minimization.
> The
> >parameter values in .mdp file are:
> 
> 
> 
> >;
> >cpp                 =  /lib/cpp
> >define              = -DPOSRES -DFLEX_SPC
>  add this                       ^^^^^^^^^^    
> 

Hi,

I'm having the same problem that Geraudis mention. I did try the define sugested
by David but I still have the following error:

Step =   218, Dmax = 2.84e-05 nm, Epot = -2.74076e+05 Fmax = 5.35193e+02
Step =   220, Dmax = 1.70e-05 nm, Epot = -2.74076e+05 Fmax = 5.09737e+02
Step =   221, Dmax = 2.04e-05 nm, Epot = -2.74076e+05 Fmax = 4.78996e+02
Step =   226, Dmax = 1.53e-06 nm, Epot = -2.74076e+05 Fmax = 4.76698e+02
Stepsize too small Converged to machine precision,
but not to the requested precision (100)

writing lowest energy coordinates.
Maximum force:  4.78996e+02

STEEPEST DESCENTS did not converge in 227 steps
  Potential Energy  = -2.74076e+05g

Conjugated Gradients did not converged the structure under 200kj. My .mdp file
to this run is:

cpp                 =  /lib/cpp
define              =  -DPOSRES -DFLEX_SPC
constraints         =  none
integrator          =  steep
nsteps              =  10000
;
;       Energy minimizing stuff
;
emtol               =  100
emstep              =  0.00001
 
nstcomm             =  1
ns_type             =  grid
rlist               =  1.5
rcoulomb            =  1.5
rvdw                =  1.5
Tcoupl              =  no
Pcoupl              =  no
gen_vel             =  no

It is possible to achieve lower gradients? 

Thanks in advance for any help,

Hugo.