[gmx-users] LIE

Alan Wilter Sousa da Silva alan at biof.ufrj.br
Wed Mar 19 14:54:50 CET 2003


Hi Dear GMX users!

	I did some tests trying to evaluate Linear Interactive Energy
method to obtain estimatives of free bind energy between protein and
ligand.

	I tried using Reaction Field, Cut-off and PME, with and without
ions or neutralising proteins (by protonation or deprotonation of some
residues).

	I did dynamics of 600 ps, 1 ns or more using Gromos96 FF, with box
at least 1.5 nm bigger than the complex and rc not greater than 1.4 nm, t
and p-coupling = berendsen.

	Evaluating only van der Waals energy (LJE) I got good results in
agreement for all cases studied.

	However Electrostatic Energy (EE) gave me very unexpected results.

	Despite the fluctuations in energy that led, in some cases,
to big RMSDs (implying unacceptable error estimation) I got in general
that EE for ligand in solvent are more negative than EE for ligand in
protein, which implies positive values for delta EE (quite weird!)
	Even considering the fact that my ligands topology could be wrong
(I tried with 4 different ligands), I tried one peptide ligand and a
dimeric protein (to calculate the EE between the two chains).
	Even for these two later cases I got strange results for EE.
Well, there was only one situation where I got negative values for delta
EE and that happen when I use R. Field (with epsilon=0 and =80) without
ions and neutralised protein.

	So:
	First, using PME under any circumstance gave EE very different
(about half) from RF and Cut-off options.  Should I not expect EE at close
values independently of the electrostatic treatment?

	Second, I would like to know if someone else has, or knows, some
published references on works using GMX and LIE to indicate me.  I must
say that I have some references using LIE (from Aquvist, Kollman etc.) but
none with GMX.

	If someone has interest in obtain more details about what was done
above, you may e-mail me directly.

Many thanks in advance for any attention. I'd appreciate very well any
commentary.

Cheers,

-----------------------
Alan Wilter S. da Silva
-----------------------
 Laboratório de Física Biológica
  Instituto de Biofísica Carlos Chagas Filho
   Universidade do Brasil/UFRJ
    Rio de Janeiro, Brasil





More information about the gromacs.org_gmx-users mailing list