AMBER Archive (2009)
Subject: [AMBER] MM-GBSA energy decomposition and pairwise decomposition question
From: Dong Xu (quantum_mania_at_yahoo.com)
We ran energy decomposition and pairwise decomposition using MM-GBSA (igb=2) on a protein-ligand complex trajectory.
The ligand is a polysacchride sugar, thus has 5 residues (each for one sugar unit) in GLYCAM naming conventions.
The energy decomposition was done between individual protein residue and the entire polysaccharides ligand. The pairwise decomposition was done between individual protein residue and individual polysacchride sugar unit.
The problem we have is that the energies from decomposition are much smaller than those from pairwise decomposition.
For example, E(GLU-ligand) < E(GLU-GAL), GAL is one of the sugar units of the ligand.
Also, if I sum the energies from energy decomposition, the total is different from the GBTOT calculated from standard MM-GBSA procedure.
Can anyone help us understand these differences?