Thanks for the reply Dr. Case
I am a little confused, maybe I have been getting lucky... but when I make
the tleap.in file, I have to specify a FF such as ff99SB or ff14SB.
I use a python script, and this is what I use
f = open("tleap.in",'w')
f.write('source oldff/leaprc.{0:s} \n'.format(ff99SB) )
f.write('source leaprc.{0:s} \n'.format(GAFF) )
f.write('MOL = loadmol2 {0:s}.mol2 \n'.format(molName) )
f.write('check MOL \n')
f.write('loadamberparams {0:s}.frcmod\n'.format(molName) )
f.write('saveoff MOL {0:s}.lib \n'.format(molName) )
f.write('saveamberparm MOL {0:s}.prmtop {1:s}.inpcrd
\n'.format(molName,molName) )
f.write('quit\n')
f.close()
Are you saying I do not need to specify ff99SB and just let GAFF do all of
the work? I was basing this on an example I saw online.
Thanks,
Melania
On Thu, Jun 6, 2019 at 9:55 AM David A Case <david.case.rutgers.edu> wrote:
> On Wed, Jun 05, 2019, Melania T wrote:
> >
> >I have been using GAFF with the ff99SB forcefield for small organic
> >molecules. Is this a reasonable setup to use?
> >
> >I know there is GAFF2, so that is an easy change... should I be using
> >something other than ff99SB?
>
> GAFF2 is recommended for small molecules. ff99SB is only for proteins,
> and should definitely be updated to ff14SB (load leaprc.protein.ff14SB
> in your tleap script.)
> >
> >The molecules are typically under 50 atoms, are mainly amines and
> ethanols.
>
> If the above is true, than you would not be using either ff99SB or
> ff14SB, since your molecules are not proteins.
>
> >
> >I am also interested in druglike molecules such as nitralin and caffeine
> >etc, these are polar molecules with many rings inside them... Will the
> same
> >methods be alright for them?
>
> Depends on your definition of "alright": GAFF2 is the best (semi-)
> automated solution that Amber offers. We have other programs (see
> especially paramfit and mdgx) for the generation of bespoke force fields
> for particular cases. And, alternatives to GAFF2 are increasingly being
> developed in other codes. Molecules with partial conjugation are
> frequently challenging to automated force field assignment protocols.
>
> ...good luck...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 Thu Jun 06 2019 - 23:00:02 PDT
