On Wed, Feb 5, 2014 at 8:22 AM, Schackert Nico <
Nico.Schackert.student.uibk.ac.at> wrote:
>
> Dear Jason,
> I am mutating a hydrogen into a methyl group in my ligand (res 236).
> One ligand is called "IIA" and the second "IIP". Water molecules (res
> 237-6904) are not involved. Thats why I don't understand the internal
> energy for res 237. Here are my two infiles.
>
> disappear atom
> &cntrl
> ntx=5, irest=1,
> ntwx=0, ntwr=20000,
> ntb=2, cut=8,
> nstlim=1000000, dt=0.001,
> temp0=300, tempi=300, ntt=3, gamma_ln=2,
> ntp=1, pres0=1, taup=2,
> icfe=1, ntf=1, ntc=1,
> clambda=0.5, nmropt=1,
> ifsc=1, idecomp=1,
> scmask=':IIA.H2',
> &end
> /
> &wt TYPE='END'
> /
> LISTOUT=POUT
> DISANG=RESTRAINTS.dat
> system
> RRES 1 6904
> END
> print
> RES 1 6904
> END
> END
>
>
> appear
> &cntrl
> ntx=5, irest=1,
> ntwx=0, ntwr=20000,
> ntb=2, cut=8,
> nstlim=1000000, dt=0.001,
> temp0=300, tempi=300, ntt=3, gamma_ln=2,
> ntp=1, pres0=1, taup=2,
> icfe=1, ntf=1, ntc=1,
> clambda=0.5, nmropt=1,
> ifsc=1, idecomp=1,
> scmask=':IIP.C11,H111,H112,H113',
> &end
> /
> &wt TYPE='END'
> /
> LISTOUT=POUT
> DISANG=RESTRAINTS.dat
> system
> RES 1 6904
> END
> print
> RES 1 6904
> END
> END
> END
>
> Can this be a technical problem?
>
If I had to guess, it is an indexing bug in the TI decomposition code. I
wouldn't be surprised if there was an off-by-one bug somewhere in the
atom/residue mapping index (so that maybe the last atom of your ligand is
assigned to the first atom of the next residue).
You can also try something that should get rid of this: turn on SHAKE
(ntf=2, ntc=2), and then add a "noshakemass=':IIP'" to your input file (or
':IIA' for the first system). This should get rid of all internal degrees
of freedom for your water molecules (and therefore eliminate the internal
contribution to dV/dl unless there really is an indexing bug as I
described).
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Feb 05 2014 - 07:00:05 PST
