thank you very much for your reply.
I have just a question on the errors, how a standard deviation of the binding energy measured can be calculated? Do I need to run a series of calculations changing the grid parameters and then calculate the SD to understand which is the influence of grid spacing and length on the obtained energy?
Is it somehow possible to understand the uncertainty of the measure related to the granularity of the grids?
Thank you infinitely,
Hi Davide --
You are correct on both counts: mg-manual is appropriate for your calculations and "bcfl focus" is the correct keyword. However, you could also use "mg-auto" if you plan to keep the coarse and fine grids' lengths and centers at the same size/location.
Nathan Baker, Ph.D.
Pacific Northwest National Laboratory
From: Davide Mercadante <firstname.lastname@example.org<mailto:email@example.com>>
Date: Wed, 26 Jan 2011 18:04:28 -0800
To: APBS List <firstname.lastname@example.org<mailto:email@example.com>>
Subject: [Apbs-users] Focusing calculations to understand individual electrostatic contributions to the binding free energy.
Dear APBS users,
I am going to perform a series of calculations using APBS to understand the contributions of salt bridges at the interface of a protein homodimer to the free energy of binding (dimerisation).
By reading the litterature I have seen that many have used a finite element method to solve PBE. They perform a refinement to a first lower resolution calculation in which the boundary points are away 20A from the molecule; the refinement step consists in a focusing calculation in which the boundary points are closer to the molecule with an increase of grid resolution.
However, reading at the APBS manual seems that a focusing calculation can be also done using a multigrid method (mg-manual) rather than of a finite element one.
I want to compute the contribution (in terms of free energy) of the salt bridges to the dimer formation by replacing the charged residues by hydrophobic isosteres.
Can you please advise if "mg-manual" suits the purposes of my study?
Do I need to prepare two input files in which the keyword bcfl has as the argument "mdh" and a second input file in which the keyword is "focus" with the new values for the boundary?
Any help will be much appreciated.
Davide Mercadante - PhD student -
Department of Chemistry
The University of Auckland
Rm 438 ext. 89171
1142 Auckland, New Zealand