AMBER Archive (2007)

Subject: Re: AMBER: total energy in sander and nmode

From: David A. Case (case_at_scripps.edu)
Date: Tue Sep 25 2007 - 09:56:49 CDT


On Tue, Sep 25, 2007, mbrut_at_laas.fr wrote:
>
> I try to do a minimization on a small molecule in explicit solvent. After
> a first minimization with SANDER, I find :
>
> NSTEP ENERGY RMS GMAX NAME NUMBER
> 500000 -3.4847E+03 3.3080E-03 4.2776E-02 O 311
>
> BOND = 276.4991 ANGLE = 10.1375 DIHED = 32.7968
> VDWAALS = 803.1828 EEL = -4420.2975 HBOND = 0.0000
> 1-4 VDW = 15.3511 1-4 EEL = -202.3743 RESTRAINT = 0.0000
>
> with the input :
>
> &cntrl
> imin = 1,
> maxcyc = 500000,
> ncyc = 500000,
> ntb = 1,
> cut = 8
> /
>
> But when I try to obtain a refinement by Newton-Raphson in NMODE, the
> total energy changes :
>
> F = -0.403742E+04 GRDMAX = 0.217857E+02 GNORM = 0.466710E+01
> E-NONB E-ELE E-HBOND E-BOND
> 0.80101E+03 -0.51059E+04 0.00000E+00 0.27650E+03
> E-ANGLE E-DIHED E-NB14 E-EEL14
> 0.10138E+02 0.32797E+02 0.15351E+02 -0.67302E+02
> E-POL E-3BOD
> 0.00000E+00 0.00000E+00
>
> with the input :
>
> &data
> ntrun = 4,
> maxcyc = 500,
> drms = 0.001,
> scee = 1.2,
> cut = 8
> /
>

The NMODE module doesn't know anything about periodic boundary conditions, nor
about PME. So, you are looking at a completely different physical system in
the two calculations: a infinite periodic system in sander, and just an
isolated unit cell in nmode.

....dac

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu