[gmx-users] Energy minimisation with different water models

[Tomek R]

This topic was already touched on the list, but I have not found any 
solution to it. I am running a simple steepest descent energy minimisation 
(em.mdp input file like in speptide tutorial) on a AAA tripeptide /OPLSAA 
forcefield, GROMACS 3.2.1 and I am getting the following results on 
different water models:

It converges well with SPC / TIP5P, but something goes very wrong with 
TIP4P. What's the problem?

Tomek

SPC:
Step=   42, Dmax= 1.6e-02 nm, Epot= -3.86495e+04 Fmax= 1.11394e+04, atom= 5
Step=   44, Dmax= 9.6e-03 nm, Epot= -3.88441e+04 Fmax= 1.82268e+03, atom= 5
.
.
Steepest Descents converged to Fmax < 2000 in 45 steps
Potential Energy  = -3.8844137e+04
Maximum force     =  1.8226837e+03 on atom 5
Norm of force     =  7.5002910e+03

TIP4P:
Step=   28, Dmax= 2.7e-06 nm, Epot= -1.39090e+08 Fmax= 2.16801e+14, atom= 
1656
Step=   30, Dmax= 1.6e-06 nm, Epot= -3.75909e+07 Fmax= 2.45911e+13, atom= 
1656
Stepsize too small, or no change in energy.
Converged to machine precision,
but not to the requested precision Fmax < 2000

Double precision normally gives you higher accuracy.
.
.
Steepest Descents converged to machine precision in 31 steps,
but did not reach the requested Fmax < 2000.
Potential Energy  = -1.3909030e+08
Maximum force     =  2.1680114e+14 on atom 1656
Norm of force     =  2.8675946e+14


TIP5P:
Step=   33, Dmax= 7.5e-03 nm, Epot= -9.64603e+03 Fmax= 4.13647e+03, atom= 35
Step=   35, Dmax= 4.5e-03 nm, Epot= -9.67248e+03 Fmax= 1.89664e+03, atom= 35
.
.
Steepest Descents converged to Fmax < 2000 in 36 steps
Potential Energy  = -9.6724756e+03
Maximum force     =  1.8966361e+03 on atom 35
Norm of force     =  2.5414328e+04

_________________________________________________________________
Want to block unwanted pop-ups? Download the free MSN Toolbar now!  
http://toolbar.msn.co.uk/