I think maybe you are performing the test with version 1.5 of
AmberTools? I mentioned that what I describe applies to AmberTools 1.4.
Indeed comparison of frcmod produced by extractFrcmod.py confirms an
inconsistency:
gly.frcmod -> multiplicity of 2 for HA2-CA-C-O
----------
H1-CT-C -O 1 0.080 180.000 -3.0 SCEE=1.2 SCNB=2.0
H1-CT-C -O 1 0.800 0.000 1.0 SCEE=1.2 SCNB=2.0
gly-nhe.frcmod -> multiplicity of 3 for HA2-CA-C-O
--------------
H1-CT-C -O 1 0.080 180.000 -3.0 SCEE=1.2 SCNB=2.0
H1-CT-C -O 1 0.000 0.000 -2.0 SCEE=1.2 SCNB=2.0
H1-CT-C -O 1 0.800 0.000 1.0 SCEE=1.2 SCNB=2.0
However I think it is preferable to compare directly the prmtop instead
of converting them to frcmod, as you cannot see the inconsistency
between HA2-CA-C-O and HA3-CA-C-O parameters in gly-nhe in the frcmod.
9-6-15-18 = HA2-CA-C-O
12-6-15-18 = HA3-CA-C-O
gly.prmtop
1 = 0.800 0.000 1.0
2 = 0.080 180.000 3.0
----------
12 6 15 18 1
12 6 -15 18 2
9 6 15 18 1
9 6 -15 18 2
gly-nhe.prmtop
3 = 0.800 0.000 1.0
4 = 0.080 180.000 3.0
5 = 0.000 0.000 2.0
--------------
12 6 15 18 3
12 6 -15 18 4
9 6 15 18 3
9 6 -15 18 5
9 6 -15 18 4
Thomas Gaillard
Le Friday 20 May 2011 à 07:16:53AM, Jason Swails a écrit :
> I still don't see the issue with those sequences. Create frcmod files using
> extractFrcmod.py and see that each dihedral (or any parameter) that is
> defined in both topology files is defined exactly the same...
>
> HTH,
> Jason
>
> On Fri, May 20, 2011 at 6:47 AM, Thomas Gaillard <
> thomas.gaillard.polytechnique.edu> wrote:
>
> > Hi Jason,
> >
> > You are right that the problem does not occur with your sequences. To
> > reproduce the problem I describe, you have to build a GLY-XXX sequence
> > and compare to isolated GLY, like this:
> >
> > source leaprc.ff99SB
> > l = sequence { GLY }
> > k = sequence { GLY NHE }
> > saveamberparm l gly.parm7 gly.rst7
> > saveamberparm k gly-nhe.parm7 gly-nhe.rst7
> > quit
> >
> > You will then see that the HA2-CA-C-O dihedral (numbered 9-6-15-18 in the
> > prmtop) has a multiplicity of 3 in gly-nhe while 2 in gly. Below are the
> > interesting parts of the prmtop.
> >
> > I think that even if the sequences are "unphysical", the assignment of
> > dihedrals by leap should be consistent.
> >
> > The version I used is the amber11-with-patches git branch so I think
> > this corresponds to AmberTools 1.4.
> >
> > gly.parm7
> > ---------
> > %FLAG ATOM_NAME
> > %FORMAT(20a4)
> > N H CA HA2 HA3 C O
> > %FLAG DIHEDRAL_FORCE_CONSTANT
> > %FORMAT(5E16.8)
> > 8.00000000E-01 8.00000000E-02 0.00000000E+00
> > %FLAG DIHEDRAL_PERIODICITY
> > %FORMAT(5E16.8)
> > 1.00000000E+00 3.00000000E+00 2.00000000E+00
> > %FLAG DIHEDRAL_PHASE
> > %FORMAT(5E16.8)
> > 0.00000000E+00 3.14159400E+00 0.00000000E+00
> > %FLAG DIHEDRALS_INC_HYDROGEN
> > %FORMAT(10I8)
> > 12 6 15 18 1 12 6 -15 18
> > 2
> > 9 6 15 18 1 9 6 -15 18
> > 2
> > 3 0 6 9 3 3 0 6 12
> > 3
> > 3 0 6 15 3
> >
> > gly-nhe.parm7
> > -------------
> > %FLAG ATOM_NAME
> > %FORMAT(20a4)
> > N H CA HA2 HA3 C O N HN1 HN2
> > %FLAG DIHEDRAL_FORCE_CONSTANT
> > %FORMAT(5E16.8)
> > 2.00000000E+00 2.50000000E+00 8.00000000E-01 8.00000000E-02
> > 0.00000000E+00
> > 4.50000000E-01 1.58000000E+00 5.50000000E-01 0.00000000E+00
> > 1.05000000E+01
> > 1.00000000E+00
> > %FLAG DIHEDRAL_PERIODICITY
> > %FORMAT(5E16.8)
> > 1.00000000E+00 2.00000000E+00 1.00000000E+00 3.00000000E+00
> > 2.00000000E+00
> > 1.00000000E+00 2.00000000E+00 3.00000000E+00 4.00000000E+00
> > 2.00000000E+00
> > 2.00000000E+00
> > %FLAG DIHEDRAL_PHASE
> > %FORMAT(5E16.8)
> > 0.00000000E+00 3.14159400E+00 0.00000000E+00 3.14159400E+00
> > 0.00000000E+00
> > 3.14159400E+00 3.14159400E+00 3.14159400E+00 0.00000000E+00
> > 3.14159400E+00
> > 3.14159400E+00
> > %FLAG DIHEDRALS_INC_HYDROGEN
> > %FORMAT(10I8)
> > 18 15 21 24 1 18 15 -21 24
> > 2
> > 18 15 21 27 1 18 15 -21 27
> > 2
> > 12 6 15 18 3 12 6 -15 18
> > 4
> > 12 6 15 21 5 9 6 15 18
> > 3
> > 9 6 -15 18 5 9 6 -15 18
> > 4
> > 9 6 15 21 5 6 15 21 24
> > 2
> > 6 15 21 27 2 3 0 6 9
> > 5
> > 3 0 6 12 5 3 0 6 15
> > 5
> > 15 24 -21 -27 11
> >
> > Thomas Gaillard
> >
> > Le Thursday 19 May 2011 à 11:21:19PM, Jason Swails a écrit :
> > > I do not see this. I used the following script:
> > >
> > > source leaprc.ff99SB
> > >
> > > l = sequence {ACE ALA GLY NHE}
> > > k = sequence {ACE ALA GLY ALA NHE}
> > >
> > > saveamberparm l nhe.parm7 nhe.rst7
> > > saveamberparm k gly.parm7 gly.rst7
> > >
> > > quit
> > >
> > > Then I used "extractFrcmod.py" to create frcmod files from nhe.parm7 and
> > > gly.parm7 (this script just extracts the parameters from these prmtop
> > files
> > > and puts it into "frcmod" format).
> > >
> > > The two frcmod files are identical (as they should be). Can you give me
> > an
> > > instance where this is not the case for you?
> > >
> > > All the best,
> > > Jason
> > >
> > > On Thu, May 19, 2011 at 8:16 PM, Thomas Gaillard <
> > > thomas.gaillard.polytechnique.edu> wrote:
> > >
> > > > When comparing prmtop files created by leap (AmberTools 1.4 version)
> > for
> > > > GLY and GLY-NHE (where GLY is GLY and not NGLY) with the ff99SB force
> > > > field, I found some inconsistencies in the dihedrals:
> > > >
> > > > The HA2-CA-C-O dihedral has a multiplicity of 3 in GLY-NHE whereas
> > > > the same dihedral in GLY has a multiplicity of 2, and the HA3-CA-C-O
> > > > dihedral in GLY-NHE has a multiplicity of 2.
> > > >
> > > > This is of course inconsistent because: 1) the HA2-CA-C-O dihedral
> > > > belongs to the GLY residue and its parameters should not be affected by
> > > > the presence of the NHE residue; 2) as HA2 and HA3 atoms are
> > equivalent,
> > > > HA2-CA-C-O and HA3-CA-C-O dihedrals should have the same parameters.
> > > >
> > > > It appears that the multiplicity of 2 for the HA2-CA-C-O dihedral
> > > > corresponds to the expected H1-CT-C-O parameters in parm99.dat:
> > > >
> > > > H1-CT-C -O 1 0.80 0.0 -1. Junmei et al, 1999
> > > > H1-CT-C -O 1 0.08 180.0 3. Junmei et al, 1999
> > > >
> > > > while the multiplicity of 3 corresponds to the addition of the
> > wild-card
> > > > parameter:
> > > >
> > > > X -C -CT-X 6 0.00 0.0 2. JCC,7,(1986),230
> > > >
> > > > I was wondering if this problem is known, and what could explain the
> > > > inconsistency in the assignment of parameters for this type of
> > dihedral.
> > > > Note that the problem also occurs with other GLY-XXX sequences.
--
Thomas Gaillard
Maître de conférences
Laboratoire de Biochimie
Ecole Polytechnique
91128 Palaiseau cedex
tel: +33 1 69 33 48 62
fax: +33 1 69 33 49 09
thomas.gaillard.polytechnique.edu
http://bioc.polytechnique.fr/~gaillard
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri May 20 2011 - 05:30:05 PDT
