On Sun, Mar 27, 2011 at 7:00 PM, case <case.biomaps.rutgers.edu> wrote:
> On Sun, Mar 27, 2011, Jason Swails wrote:
> >
> > I decided to throw a little (not much) time at validating ff10 wrt what's
> > written in the manual. I've written a little script to do this (attached
> > here).
> >
>
> Thanks!!
>
> > As expected, the only ones that match (and match exactly through 100
> steps)
> > is the ff10/ff99SB for proteins (good news, no?).
>
> Yes, confirming tests made here by InSuk Joung.
>
> >
> > For the other ones, I don't know what frcmods to load on top of ff99bsc0
> in
> > order to get what's "expected" for ff10.
>
> According to the supplementary material from Banas10, you should do this:
>
> source $AMBERHOME/dat/leap/cmd/leaprc.ff99bsc0
> loadamberprep all_nuc94_ol_bsc0.in
> loadamberparams frcmod.ol.dat
>
> The files for the last two lines are in amber11/dat/leap/ff99bsc0chiOL.
> Setups created this way should then be tested against leaprc.ff10.
>
I don't have this folder in my repo. Where can I find them?
> I'm really, really hoping that tec will chime in, since I think(?) he has
> access to the "real" prmtops used for the tetraloop simulations reported
>
If he has the prmtop, then I've added a cute little python script to the
Amber bin directory "extractFrcmod.py" that will take a prmtop file and back
out an frcmod from it:
extractFrcmod.py <prmtop> <frcmod>
> in the Banas10 paper. In addition, we want to be sure that we are really
> doing the "right thing" with DNA as well. (A DNA setup with ff10 should be
> identical to one from ff99bsc0.)
>
The DNA setup for ff10 has 2 differences from ff99bsc0; the OS-CT-N*-CK
dihedral term. According to my frcmod utility:
ff10:
OS-CT-N*-CK 1 1.065 6.229 -2.0 SCEE=1.2 SCNB=2.0
OS-CT-N*-CK 1 0.443 168.650 -3.0 SCEE=1.2 SCNB=2.0
OS-CT-N*-CK 1 0.256 3.975 -4.0 SCEE=1.2 SCNB=2.0
OS-CT-N*-CK 1 0.705 74.756 1.0 SCEE=1.2 SCNB=2.0
ff99bsc0:
OS-CT-N*-CK 1 0.000 0.000 -2.0 SCEE=1.2 SCNB=2.0
OS-CT-N*-CK 1 2.500 0.000 1.0 SCEE=1.2 SCNB=2.0
I think this is problematic in parm10.dat. Lines 671-674 read:
OS-CT-N*-CK 1 0.70510 74.7558 -1. ol3 chi gua
OS-CT-N*-CK 1 1.06546 6.2286 -2. ol3 chi gua
OS-CT-N*-CK 1 0.44273 168.6503 -3. ol3 chi gua
OS-CT-N*-CK 1 0.25602 3.9746 4. ol3 chi gua
Whereas lines 761-762 read:
OS-CT-N*-CK 1 0.00 000.0 -2. parm98,
TC,PC,PAK
OS-CT-N*-CK 1 2.50 0.0 1. parm98,
TC,PC,PAK
It would seem as though ff10 doesn't know how to make up its mind (although
it's nice to see that my frcmod creator works :) ).
(Closer inspection of the frcmods created by the script I attached to the
last email shows that this issue also exists with the dihedral OS-CT-N*-CM,
which is the same as the one above in ff99bsc0, but slightly different from
the one above for ff10).
This just about hits the limit of how useful I can be here. It seems like
we need to address some issues with the nucleic acid force fields. If
someone gives me the aforementioned RNA force field files to attach to
ff99bsc0 to attack RNA with, I'll add it to the script and report back on
what I find.
Hope this helps,
Jason
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Mon Mar 28 2011 - 01:00:02 PDT
