AMBER: Question about partial charge for ligand molecules

From: Woojin Lee <>
Date: Thu, 29 Dec 2005 11:37:47 +0900

Hi folks,

   I generated ligand molecules with MMFFs94 f.f. in Macromodel for binding
affinity MD simulation with Amber, as there is
No crystal structure for lignad-protein complex.
   For consistency, we need to RESP charge. From the structures, we
performed HF 6-31G* single point calculations
with Gamess program to get RESP charges "without geometry optimization" by
the same level of theory and basis sets.

   We examined the am1-bcc by divcon and Antechamber, but we did not find a
way to assign equal charges to symmetric
atoms. So we decided to get partial charge for ligands by RESP protocol.

   What I worry about is that the structure used in single point energy is
not the one from HF-optimization.

what I expect is:
1. MM minimized structures have reasonable geometries.
2. Even we perform HF optimzation from MM-minimized one, it will not alter
the geometry much.
3. Even internal geometry changed by some extent, charge will not change

In addition, the non-parallel Gamess took too much time to run a
optimization job for current ligands( ca. 60 atoms and more...)
(our cluster: Xeon 2.8 GHz RH9)

   Q1) Is there any input keyword or method to assign eual charges for
symmetric atoms by am1-bcc protocol?
   Q2) Is the procedure of ESP generation not from HF-optimized geom, but
from MM-minimized geom can be valid?

Thanks in advance

Department of Chemistry
Seoul National University

Woojin Lee

The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Wed Jan 04 2006 - 18:17:02 PST
Custom Search