Re: how to constrain a dihedral angle?

From: Leonardo G. Trabuco (ltrabuco_at_ks.uiuc.edu)
Date: Sat Mar 25 2006 - 17:48:57 CST

On Sat, Mar 25, 2006 at 10:17:29PM +0800, xiaojing gong wrote:
> Thank you for your advice, if I want to constrain the dihedral
> CB--CG--CD--NE of Arg residue. Do I need to define a new bond ?
> would you give me more details about the technique . I am a fresh man
> in this field, I really need someone to give me more details.

Ok, this is going to be a loOong e-mail... :-)

First of all, if you're not familiar with CHARMM topology and parameter
files, read the corresponding appendices in the NAMD tutorial:
http://www.ks.uiuc.edu/Training/Tutorials/

You want to constrain the dihedral CB-CG-CD-NE, which would correspond
to CT2-CT2-CT2-NC2 in the parameter file (e.g. par_all27_prot_na.prm),
but if won't find this entry there if you try. But there is an entry
like this:

X CT2 CT2 X 0.1950 3 0.00 ! ALLOW ALI
                ! alkane update, adm jr., 3/2/92

The 'X' is not really an atom type, it's a wild card. In this case, when
NAMD will use these parameters for CT2-CT2-CT2-NC2. If there were an
entry for CT2-CT2-CT2-NC2, NAMD would use it instead, even though there
is a wild card somewhere.

Now, if you look at the topology file (e.g. top_all27_prot_na.top), you
will see that the only occurence of the dihedral CT2-CT2-CT2-NC2 is in
arginine. So, the quick and dirty way would be to define a new entry in
the parameter file for the dihedral CT2-CT2-CT2-NC2, using the same
parameters as above but with a higher force constant:

CT2 CT2 CT2 NC2 10.0000 3 0.00 ! same as X CT2 CT2 X,
                ! with a higher force constant (10.0 is just an example)

I'm not sure if the order matters, but if you define the "specific"
dihedral before the "wild card", you will be safe. I.e., you could copy
this entry in the beginning of the DIHEDRALS section in the parameter
file.

The problem with this approach is the following. If you have more than
one Arg residue in your protein, this will apply the constrain to them
all. If that's not what you want, than the solution is not that simple.

So here's what I would do. I would define a new residue in the topology
file, coping everything in the RESI ARG to RESI ARGC, say. Now I'd alter
the atom types for CB, CG, CD, and NE (from CT2 and NC2 to CT2C and NC2,
for example). Now the topology of the new residue ARGC looks like this:

-----------------------------------------------------------------------

RESI ARGC 1.00
GROUP
ATOM N NH1 -0.47 ! | HH11
ATOM HN H 0.31 ! HN-N |
ATOM CA CT1 0.07 ! | HB1 HG1 HD1 HE NH1-HH12
ATOM HA HB 0.09 ! | | | | | //(+)
GROUP ! HA-CA--CB--CG--CD--NE--CZ
ATOM CB CT2C -0.18 ! | | | | \
ATOM HB1 HA 0.09 ! | HB2 HG2 HD2 NH2-HH22
ATOM HB2 HA 0.09 ! O=C |
GROUP ! | HH21
ATOM CG CT2C -0.18
ATOM HG1 HA 0.09
ATOM HG2 HA 0.09
GROUP
ATOM CD CT2C 0.20
ATOM HD1 HA 0.09
ATOM HD2 HA 0.09
ATOM NE NC2C -0.70
ATOM HE HC 0.44
ATOM CZ C 0.64
ATOM NH1 NC2 -0.80
ATOM HH11 HC 0.46
ATOM HH12 HC 0.46
ATOM NH2 NC2 -0.80
ATOM HH21 HC 0.46
ATOM HH22 HC 0.46
GROUP
ATOM C C 0.51
ATOM O O -0.51

[the rest is the same as for ARG]

-----------------------------------------------------------------------

Besides, you have to define the MASS entries for the two new atom types
in the beginning of the topology file, just coping from CT2 and NC2:

-----------------------------------------------------------------------
MASS 200 CT2C 12.01100 C ! same as CT2
MASS 201 NC2C 14.00700 N ! same as NC2
-----------------------------------------------------------------------

Now, we can define a new torsion angle CT2C-CT2C-CT2C-NC2C, the same way
as above, that will only be applied to ARGC, and not ARG. But now all
parameters that relate to CT2C or NC2C are not present in the parameter
file... :-(

What we have to do now is to define new entries for every parameter that
is not found anymore because we changed the atom type, coping them from
the corresponding entries with CT2 and NC2. For this, I recommend you to
write down all the vDW, bonds, angles, dihedrals, and impropers that
involve CB, CG, CD, and NE. Now look at the parameter file for every
single parameter that you've listed, and copy it to a new entry, changing
the atom type to your new atom type (CT2C and NC2C). The result is the
following (if I didn't make any mistake, check it!):

-----------------------------------------------------------------------

!BONDS FOR ARGC
CT2C CT2C 222.500 1.5300 ! same as CT2 CT2
CT2C CT1 222.500 1.5380 ! same as CT2 CT1
HA CT2C 309.000 1.1110 ! same as HA CT2
NC2C CT2C 261.000 1.4900 ! same as NC2 CT2
NC2C HC 455.000 1.0000 ! same as NC2 HC
NC2C C 463.000 1.3650 ! same as NC2 C

!ANGLES FOR ARGC
CT2C CT2C CT2C 58.350 113.60 11.16 2.56100 ! same as CT2 CT2 CT2
HA CT2C CT1 33.430 110.10 22.53 2.17900 ! same as HA CT2 CT1
CT2C CT2C CT1 58.350 113.50 11.16 2.56100 ! same as CT2 CT2 CT1
HB CT1 CT2C 35.000 111.0000 ! same as HB CT1 CT2
NH1 CT1 CT2C 70.000 113.5000 ! same as NH1 CT1 CT2
CT2C CT1 C 52.000 108.0000 ! same as CT2 CT1 C
CT2C NC2C C 62.300 120.0000 ! same as CT2 NC2 C
HC NC2C CT2C 40.400 120.0000 ! same as HC NC2 CT2
NC2C CT2C CT2C 67.700 107.5000 ! same as NC2 CT2 CT2
NC2C CT2C HA 51.500 107.5000 ! same as NC2 CT2 HA
HC NC2C C 49.000 120.0000 ! same as HC NC2 C
NC2C C NC2 52.000 120.00 90.00 2.36420 ! same as NC2 C NC2
HA CT2C HA 35.500 109.00 5.40 1.80200 ! same as HA CT2 HA
HA CT2C CT2C 26.500 110.10 22.53 2.17900 ! same as HA CT2 CT2

!DIHEDRALS FOR ARGC
CT2C CT2C CT2C NC2C 0.1950 3 0.00 ! increase this bond constant
                ! to constrain the dihedral CB-CG-CD-NE in ARGC
H NH1 CT1 CT2C 0.0000 1 0.00 ! same as H NH1 CT1 CT2
O C CT1 CT2C 1.4000 1 0.00 ! same as O C CT1 CT2
X CT1 CT2C X 0.2000 3 0.00 ! same as X CT1 CT2 X
X CT2C CT2C X 0.1950 3 0.00 ! same as X CT2 CT2 X
X CT2C NC2C X 0.0000 6 180.00 ! same as X CT2 NC2 X
X C NC2C X 2.2500 2 180.00 ! same as X C NC2 X
CT2C CT1 NH1 C 1.8000 1 0.00 ! same as CT2 CT1 NH1 C
NH1 C CT1 CT2C 0.0000 1 0.00 ! same as NH1 C CT1 CT2

!IMPROPERS FOR ARGC
NC2C X X C 40.0000 0 0.0000 ! same as NC2 X X C

!Lennard-Jones parameters for ARGC
CT2C 0.000000 -0.055000 2.175000 0.000000 -0.010000 1.900000 ! same as CT2
NC2C 0.000000 -0.200000 1.850000 ! same as NC2
                ! JG 8/27/89; note: NH1 in ARG was changed to NC2.

-----------------------------------------------------------------------

You can increase the force constant of the CT2C CT2C CT2C NC2C dihedral
to constrain it as much as you want.

Now you can use the 'mutate' command of psfgen to change the specific
Arg residues that you want to constrain into ARGC.

I think that'll do the trick. Please let me know if that works for you.

Good luck! :-)
Leo

--
Leonardo Giantini Trabuco
Ph.D. student
Center for Biophysics and Computational Biology
University of Illinois at Urbana-Champaign
So it doesn't mess up the flow of reading.
> How come?
> > I prefer to reply inline.
> > > What do you do instead?
> > > > No.
> > > > > Do you like top-posting?

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:43:26 CST