Hi,
I have two different energies from mme and mme2 routines of NAB. Here is
the nab code:
---------------------------------------
molecule m;
int num, ier, i;
float m_xyz[dynamic], f_xyz[dynamic];
float dgrad, fret;
m = getpdb( argv[2] );
allocate m_xyz[3*m.natoms];
allocate f_xyz[3*m.natoms];
//allocate h_xyz[3*m.natoms];
//allocate mass[m.natoms];
readparm(m, argv[3]);
//setxyz
for ( i = 1; i <= m.natoms*3; i++ ) {
m_xyz[i] = atof( argv[i+3] );
}
// Energy
mm_options("cut=999.0, ntpr=1, nsnb=999, gb=0");
mme_init(m, NULL, "::ZZZZ", m_xyz, NULL);
fret = mme( m_xyz, f_xyz, -1 );
printf( "mme returns %.60e\n", fret );
// get the normal modes:
num=3*m.natoms-6;
ier = nmode( m_xyz, 3*m.natoms,mme2, num ,0,0.0,0.0,0);
printf("nmode returns %d\n", ier );
---------------------------------------
This is the output of the energy evaluation:
-- Parameter topology includes 10-12 terms:
These are assumed to be zero here (e.g. from TIP3P water)
mm_options: cut=999.0
mm_options: ntpr=1
mm_options: nsnb=999
mm_options: gb=0
bond: 5.455061285
angle: 0.531769495
dihedral: 0.204893077
enb14: -0.500399875
eel14: -1576.959704503
enb: 14.870758041
eel: 984.019892601
egb: 0.000000000
econs: 0.000000000
esurf: 0.000000000
Total: -572.377729879
frms: 6.712036879
mme returns
-5.723777298788290863740257918834686279296875000000000000000000e+02
iter Total bad vdW elect nonpolar genBorn
frms
ff: 1 -571.49 7.08 14.37 -592.94 0.00 0.00
6.90e+00
Energy = -5.7148548010e+02
RMS gradient = 6.8996841060e+00
dysev time = 0.04 seconds
--<snip>--
nmode analysis shows the energy evaluation at the beginning but the number
is different. Electrostatic and vdW energies seem to be the same but bonded
energies are different. I don't know which part of the energy makes the
difference. Could anybody tell me how to printout detailed decomposed
energies using mme2 routine? It looks like mme2 takes more than 5 arguments
which are more than the arguments described in AmberTools manual.
Best,
InSuk Joung
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Sat Aug 14 2010 - 19:30:03 PDT
