[gmx-users] RE:g_lie

[lloyd riggs]

Message: 1
Date: Sun, 11 Sep 2011 13:01:28 +0200
From: <gal.fradin at live.biu.ac.il>
Subject: [gmx-users] g_lie
To: <gmx-users at gromacs.org>
Message-ID: <BLU160-W601D2AB913084C46676AFEE9030 at phx.gbl>
Content-Type: text/plain; charset="windows-1255"


Greetings to all!
 
I have a question for the g_lie function users...
 
Is there anyone who used this function successfully including the -Elj and -Eqq options???
 
When I'm trying to use this function with the above mentioned options I'm getting Zeros for both Elj and Eqq
 
or any other value which I set using the command... for instance if I set the Elj to be 1 then the results are 'ones'...
 
It seems like a bug... 
 
I'm using the 4.5.3 version...
 
 
thanks in advance,
 
Gal 

I have.  It crankes out good values for small molecules <30 atoms, but is way off for large molecules (ie proteins of 100 amino acids or more).  It suggest in the paper and gromacs slidshow that the alpha and beta, or perhaps a gamma function need to be determined for 100+ protein protein interactions, such as was done to achieve the averidged values used for the small molecules, to be determine properly for the electrostatic and coulombic contributions. 

You have to determine the Elj or Eqq first using g_energy, or another analysis.  From what I can tell, it uses the averidges over a run for state B or A and then just does the Math function described in the paper or gromacs manual.
-- 
NEU: FreePhone - 0ct/min Handyspartarif mit Geld-zurück-Garantie!		
Jetzt informieren: http://www.gmx.net/de/go/freephone