Re: AMBER: hessian calculation

From: <sebnem.mercury.chem.pitt.edu>
Date: Tue, 28 Mar 2006 19:20:18 -0500 (EST)

I actually don't need different force constants for restraints. By chance
when i was playing with the force constant , i noticed this. After that i
become suspicious about my calculation. There should not be any
contribution of this force constant on the part of the hessian from
non-restrained atoms.
Here is an example of two inputs that i used with same coordinate,topology
and reference coordinates for restraints.

obtain gas-phase normal modes
 &data
     ntrun = 1, nsave=1, ndiag=2, cut=999.0, ntx=1,
     nprint=1,
     drms = 5, maxcyc=1,
     scnb=2.0, scee=2.0,idiel=1,
     nvect=1,ntx=1,ntxo=1,icons=1,
 &end
group
 1
RES 35 279
END
END

 obtain gas-phase normal modes
 &data
     ntrun = 1, nsave=1, ndiag=2, cut=999.0, ntx=1,
     nprint=1,
     drms = 5, maxcyc=1,
     scnb=2.0, scee=2.0,idiel=1,
     nvect=1,ntx=1,ntxo=1,icons=1,
 &end
group
 10
RES 35 279
END
END


For hessian calculation, i used nmode.f .Inside subroutine fillm, I put a
print statement inside the loop. I suppose the first value
that i will get from this will be a(1,1) which does not have anything to
the with the force constant on residues 35-279. However i got two
different values for a(1,1) as well as some other elements.
Am I doing something wrong?
thanks again,
sebnem

On Tue, 28 Mar 2006, David A. Case wrote:

> On Tue, Mar 28, 2006, sebnem wrote:
>
> > My question is how the force constant on the restrained atoms
> > can effect the second derivative of the potential with respect to the
> > non-restrained coordinates?
>
> It should not have such an effect. I wonder if your qualitative description
> really captures exactly what you did. Why were you changing the force
> constant for the restraining forces?
>
> ...dac
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
>
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun Apr 02 2006 - 06:10:09 PDT
Custom Search