AMBER Archive (2009)

Subject: Re: [AMBER] RE: Listing gaff parameters in frcmod

Date: Fri May 08 2009 - 09:23:45 CDT


if you really want to have the whole gaff forcefield as a dummy
forcefield, you could just set all the vdw parameters at the end of
gaff.dat to zero and load leaprc.gaff (or better, do this with a copy of
the gaff files of course...). If you do organic molecules in e.g. TIP3P
water, that should work. For the initial state you would use the original
gaff FF, for the final (dummy) state the modified one.

If you are doing lots of complete molecule disappearances, I would however
advocate switching to amber10 and using the soft core scaling algorithm
that offers improved accuracy for precisely this type of calculation.
Please see:

  author = {Steinbrecher, T. and Mobley, D.L. and Case, D.A.},
  title = {{Non-linear scaling schemes for Lennard-Jones interactions in
free energy calculations}},
  journal = {J. Chem. Phys.},
  volume = {127},
  pages = {214108},
  year = {2007}

Linear scaling wont work for this type of calculations and the paper
outlines a few problems you will run into when using klambda-scaling.
Using soft core would also save you the work of building all the dummy and
chargeless intermediate prmtop files.

Kind Regards,

Dr. Thomas Steinbrecher
BioMaps Institute
Rutgers University
610 Taylor Rd.
Piscataway, NJ 08854

AMBER mailing list