Dear all,
I am currently having an issue with imaging using ptraj;
If I calculate the single point energy of a structure before imaging and
centering with ptraj and after, the energies differ, this also affects
the binding energies of the system (the example below is for ethane and
water but this was also true for phenol bound to the T4 lysozyme double
mutant L99A/M102Q).
Test case: 1 ethane and 1 water molecule (TIP3P)
The original pdb coordinates:
ATOM 1 C1 ETH X 1 12.354 14.419 4.092
ATOM 2 H11 ETH X 1 13.403 14.249 4.345
ATOM 3 H12 ETH X 1 11.785 13.586 4.510
ATOM 4 H13 ETH X 1 11.889 15.277 4.581
ATOM 5 C2 ETH X 1 12.267 14.370 2.502
ATOM 6 H21 ETH X 1 12.877 15.118 1.991
ATOM 7 H22 ETH X 1 11.230 14.555 2.217
ATOM 8 H23 ETH X 1 12.564 13.428 2.038
TER
ATOM 384 O WAT X 127 11.898 11.485 0.453
ATOM 385 H1 WAT X 127 12.553 10.871 0.784
ATOM 386 H2 WAT X 127 11.971 11.422 -0.500
END
and energies:
NSTEP ENERGY RMS GMAX NAME NUMBER
1 5.0137E+00 9.8262E+00 3.3185E+01 C1 1
BOND = 1.0251 ANGLE = 3.0366 DIHED = 0.0888
VDWAALS = 0.0775 EEL = -0.1070 HBOND = 0.0000
1-4 VDW = 0.0101 1-4 EEL = 0.8826 RESTRAINT = 0.0000
The coordinates after ptraj imaging the ethane to the center of the box:
ATOM 1 C1 ETH 1 0.058 0.044 0.808
ATOM 2 H11 ETH 1 1.107 -0.126 1.061
ATOM 3 H12 ETH 1 -0.511 -0.789 1.226
ATOM 4 H13 ETH 1 -0.407 0.902 1.297
ATOM 5 C2 ETH 1 -0.029 -0.005 -0.782
ATOM 6 H21 ETH 1 0.581 0.743 -1.293
ATOM 7 H22 ETH 1 -1.066 0.180 -1.067
ATOM 8 H23 ETH 1 0.268 -0.947 -1.246
TER 0 0
ATOM 9 O WAT 2 -0.398 -2.890 -2.832
ATOM 10 H1 WAT 2 0.257 -3.504 -2.500
ATOM 11 H2 WAT 2 -0.325 -2.953 -3.785
and the energies after using ptraj:
NSTEP ENERGY RMS GMAX NAME NUMBER
1 5.0126E+00 9.8296E+00 3.3185E+01 C1 1
BOND = 1.0254 ANGLE = 3.0366 DIHED = 0.0888
VDWAALS = 0.0760 EEL = -0.1070 HBOND = 0.0000
1-4 VDW = 0.0101 1-4 EEL = 0.8826 RESTRAINT = 0.0000
This is a difference of 0.0011 kcal/mol in the energies, which when
calculating the interaction energies of this system leads to an error of
0.0014 kcal/mol.
This error scales with the number of atoms within the system, the
protein-ligand system tested had an error in the interaction energies of
0.2 kcal/mol.
If I move the ethane to the center of the box not using ptraj I get the
following coordinates:
ATOM 1 C1 ETH 1 0.058 0.044 0.808
ATOM 2 H11 ETH 1 1.107 -0.126 1.061
ATOM 3 H12 ETH 1 -0.511 -0.789 1.226
ATOM 4 H13 ETH 1 -0.407 0.902 1.297
ATOM 5 C2 ETH 1 -0.029 -0.005 -0.782
ATOM 6 H21 ETH 1 0.581 0.743 -1.293
ATOM 7 H22 ETH 1 -1.066 0.180 -1.067
ATOM 8 H23 ETH 1 0.268 -0.947 -1.246
TER
ATOM 384 O WAT 127 -0.398 -2.890 -2.831
ATOM 385 H1 WAT 127 0.257 -3.504 -2.500
ATOM 386 H2 WAT 127 -0.325 -2.953 -3.784
END
which, although difficult to see the 3rd decimal place in the z-axis of
O and H2 differ from the coordinates of ptraj. These coordinates,
produce the same energies as the original (unptraj'd) version:
NSTEP ENERGY RMS GMAX NAME NUMBER
1 5.0137E+00 9.8262E+00 3.3185E+01 C1 1
BOND = 1.0251 ANGLE = 3.0366 DIHED = 0.0888
VDWAALS = 0.0775 EEL = -0.1070 HBOND = 0.0000
1-4 VDW = 0.0101 1-4 EEL = 0.8826 RESTRAINT = 0.0000
I was wondering if anyone else has noticed this discrepancy?
Thanks for your time,
Chris Sampson
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Jul 12 2012 - 03:00:02 PDT
