[Apbs-users] binding energies and grid size

[<ki...@be...>]

Hello -

I am a new APBS user, and I would like to calculate the binding free
energy for a protein-protein complex, and several mutants of that complex.
I tried doing the calculations on my complex using three different grid
sizes and obtained three different binding energies. All other parameters
besides grid dimension were identical between the calculations. When I
calculate the complex binding energies I calculate:

complex energy - mol 1 energy - mol 2 energy

Where all grid dimension/length/centers are the same for the complex, mol1
and mol2 - and thus I thought I was eliminating the "self energy"
components/reducing the effects of grid size artifacts?? Am I getting
different energies merely because the calculation is more accurate at the
smaller grid size or is there some other difference I need to be
accounting for?  Below are my results for the different grid sizes.
Many thanks for your help with this!!
Kim Reynolds

run 1:
dimension: 97 x 97 x 97
cglen: 102.1989 118.7246 88.4272
fglen: 80.1170 89.8380 72.0160
complex binding energy: 4.87 E+02

run 2:
dimension: 129 x 129 x 129
cglen: 102.1989 118.7246 88.4272
fglen: 80.1170 89.8380 72.0160
complex binding energy: 3.11 E+02

run 3:
dimension: 161 x 161 x 161
cglen: 102.1989 118.7246 88.4272
fglen: 80.1170 89.8380 72.0160
complex binding energy: 4.405 E+02