Can you please provide some examples of the system you’re testing and the input files you’re generating?
Pacific Northwest National Laboratory
From: LI An-Bang [mailto:email@example.com]
Sent: Thursday, June 14, 2012 6:53 AM
Subject: [apbs-users] calculated energies by APBS in auto-mode
As we know, APBS can work in two modes: auto-mode, in which a coarse grid and a fine grid are used, and manual-mode, in which one grid is used. I am interested in the convergence of APBS, so I run APBS in auto-mode (the input files were generated by PDB2PQR) and in manual-mode at grid spacing 1.1A, 1.0A, 0.9A, 0.8A, 0.7A, 0.6A, 0.5A, 0.4A, 0.3A and 0.2A (the input files were generated by a homemade Perl script, the grid spacing was assign by keyword "grid") to calculate the electrostatic solvation energies of some protein structures. I found the calculated energies of auto-mode are close to that of manual-mode at 0.5A instead of anticipated 0.2A. And there are differences as large as 20-80kcal/mol between results at 0.5A and 0.2A.
It is generally regarded that the calculated results at fine grid spacing are converged ones. It seems to me that the calculated energies by APBS in auto-mode are not the converged results. Am I right?
Physics Department, HuaZhong Normal University, China