Hello
My message below describes my attempts to get 10_12 potential working. I
thought I got it working for the H nonbonds, but it is not. I can tell
because the vdw energies are too high. The only component of the vdw that
works relates to the one where the ICO index is positive.
Andrew Petersen
---------------------------- Original Message ----------------------------
Subject: 10_12 potential
From: aapeters.ncsu.edu
Date: Mon, February 8, 2010 3:14 am
To: amber.ambermd.org
--------------------------------------------------------------------------
Hello
I'm trying to use the 10-12 potential with Langevin dynamics with serial
sander in Amber9.
I found the special installation instructions in the manual
(cd $AMBERHOME/src/sander, make clean, make AMBERBUILDFLAGS'-DHAS_10_12'
sander)
Also there was some discussion in the Amber Reflector archive dated Feb.
28th 2008, that alluded to more modifications. My understanding is that
these additional instructions were for a specific case of using "oldstyle
prmtop" and requiring large arrays. I'm using the new prmtop style, with
no array limit problems.
Am I right in thinking that the simple installation instructions in the
manual is all I need?
I created a 10-12 prm file by manually editing one of my 12-6 parameter
files.
The ICO indices (%FLAG NONBONDED_PARM_INDEX) are all negative (I want the
10_12 potential for all the atoms, C and H, of my methane system).
The A and B vdw numbers are placed in the %FLAG HBOND_ACOEF and %FLAG
HBOND_BCOEF positions in the prm file.
If the H...H, H...C and C...H related indices are negative, but the C...C
related index stays +ve, all seems to work. But if I make the forth index
negative, for the C...C interaction, all my VDW energies are zero, and my
molecules crash into each other.
Is there a way I can make the C...C interaction 10_12 too?
Also, I need to instruct the cluster manager to re-compile parallel sander
in Amber10. Am I correct that the simple instructions in the manual is
all that is needed?
Thank you
Andrew Petersen
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Feb 08 2010 - 03:30:02 PST
