Re: [AMBER] Minimization in sander

From: Anglea A. <a.anglea90.yahoo.com>
Date: Wed, 2 Nov 2011 05:22:01 -0700 (PDT)

Dear Professor David
Thanks for answering my email.
I missed writing 10^5 for energy values.
The energies of the calculated steps using LBFGS method:
NSTEP    1            ENERGY        -5.7810E+05
NSTEP     27        ENERGY         -5.2950E+05

The NSTEP is the final result.
The input of my job is:
 &cntrl

 imin   = 1,
 maxcyc = 100000,
 drms   = 1.0E-5,
 ntb    = 0,
 ntmin = 3,
 cut    = 15.0,
 ntb=0,
 igb=0,
/
 &lmod
 xmin_method = "LBFGS"

/

Is that normal?
as well, using TNCG method, gives unrealistic optimized structure of the complex, where the ligand pushed out the active site. 
In addition, the final energy of the TNCG is higher than LBFGS. I read that TNCG is better than LBFGS, isn't it?
Cheers;

----- Original Message -----
From: David A Case <case.biomaps.rutgers.edu>
To: Anglea A. <a.anglea90.yahoo.com>; AMBER Mailing List <amber.ambermd.org>
Cc:
Sent: Wednesday, November 2, 2011 11:47 AM
Subject: Re: [AMBER] Minimization in sander

On Wed, Nov 02, 2011, Anglea A. wrote:

> I used LBFGS method for sander minimization. The initial energy of the
> complex is -5.78 and the final energy is -5.29 kcal.mol-1. why the final
> energy is higher than the initial energy?

We'd have to see the details.  Minimizations are not guaranteed to be
monotonically decreasing.  What caused the minimization to stop?  Was it
because it had reached the requested maximum number of cycles?  Also, having a
protein/ligand total energy be so close to zero is rather unusual.  I think
we would have to see your inputs and outputs to be of much help.

...dac


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Nov 02 2011 - 05:30:05 PDT
Custom Search