Jennifer Williams wrote:
Thanks again for the help. I?ve given it a go but am not overly
confident or exactly sure how I would translate this method to my system.
This is because rather than having a chain with a well defined start and
finish I have a giant covalent structure (like a web) where each silicon
is tetrahedrally bound to oxygen (as in quartz).
O?
|
?O-Si-O ?
|
O?
Here I describe my efforts so far.
I have defined a monomer (my internal unit) as an SiO4 tetrahedra.
Therefore each monomer would have to form 4 bonds with other monomers. I
have defined my internal residue like this:
; Internal residue
[ MCM_I ]
[ atoms ]
SI SI 1.280 1
O1 O1 -0.640 1
O2 O2 -0.640 1
O3 O3 -0.640 1
O4 O4 -0.640 1
[ bonds ]
SI O1
SI O2
SI O3
SI O4
O1 -SI
O2 -SI
O3 +SI
O4 +SI
As an aside-This means that each residue is not neutral as the charges
cancel out over the entire molecule and not over a single residue-I am
not sure of the implications of this.
To complicate matters, in my structure not all of the oxygens are bonded
oxygens (i.e where each O is bonded to 2 silicons, some of the oxygens
terminate in hydroxyl groups). This means that I have will have 3 types
of terminal/starting chain
1. Si, O, O, OH
2. SI, O, OH, OH
3. SI, OH, OH, OH (the group which really does terminate)
Here are my terminal and starting residues:
; terminal residue 1 (3OH groups)
[ MCM_T1 ]
[ atoms ]
SI SI 1.280 1
OH1 OH1 -0.502 1
H1 H1 0.206 1
OH2 OH2 -0.502 1
H2 H2 0.206 1
OH3 OH3 -0.502 1
H3 H3 0.206 1
O4 O4 -0.640 1
[ bonds ]
SI OH1
SI OH2
SI OH3
SI O4
OH1 H1
OH2 H2
OH3 H3
O4 -SI
; terminal residue 2 (2 OH groups)
[ MCM_T2 ]
[ atoms ]
SI SI 1.280 1
OH1 OH1 -0.502 1
H1 H1 0.206 1
OH2 OH2 -0.502 1
H2 H2 0.206 1
O3 O3 -0.640 1
O4 O4 -0.640 1
[ bonds ]
SI OH1
SI OH2
SI O3
SI O4
OH1 H1
OH2 H2
O3 -SI
O4 -SI
; terminal residue 3 (1 OH group)
[ MCM_T3 ]
[ atoms ]
SI SI 1.280 1
OH1 OH1 -0.502 1
H1 H1 0.206 1
O2 O2 -0.640 1
O3 O3 -0.640 1
O4 O4 -0.640 1
[ bonds ]
SI OH1
SI O2
SI O3
SI O4
OH1 H1
O2 -SI
O3 -SI
O4 -SI
As each of these groups could equally be starting groups-I have defined
them as such by changing the minus sign to a plus
; starting residue 1
[ MCM_S1 ]
[ atoms ]
SI SI 1.280 1
OH1 OH1 -0.502 1
H1 H1 0.206 1
OH2 OH2 -0.502 1
H2 H2 0.206 1
OH3 OH3 -0.502 1
H3 H3 0.206 1
O4 O4 -0.640 1
[ bonds ]
SI OH1
SI OH2
SI OH3
SI O4
O4 +SI
; starting residue 2
[ MCM_T2 ]
[ atoms ]
SI SI 1.280 1
OH1 OH1 -0.502 1
H1 H1 0.206 1
OH2 OH2 -0.502 1
H2 H2 0.206 1
O3 O3 -0.640 1
O4 O4 -0.640 1
[ bonds ]
SI OH1
SI OH2
SI O3
SI O4
O3 +SI
O4 +SI
; starting residue 3
[ MCM_T3 ]
[ atoms ]
SI SI 1.280 1
OH1 OH1 -0.502 1
H1 H1 0.206 1
O2 O2 -0.640 1
O3 O3 -0.640 1
O4 O4 -0.640 1
[ bonds ]
SI OH1
SI O2
SI O3
SI O4
O2 +SI
O3 +SI
O4 +SI
There are a few problems with this:
1. I don?t know how to go about splitting my large .pdb file into
monomers. At the moment it is ordered by atomtype VMD doesn?t recognise
my self- defined SiO2 tetrahedra as monomers so I can?t sort using that.
There is no way I can do this manually by looking at the coordinates.
2. Looking at the terminal residue 1 for example, I have defined the
only non-bonded oxygen as O4-however it could equally be O1, O2 or
O3-this leads to a number of possible combinations of my terminal and
internal residues.
3. There is in fact no such thing as a terminal residue (except in
the case of Terminal residue 1 which is rare). It is more common to have
a 2 OH groups on a silicon meaning the other oxygens bond to further
residues.
I can see how this method works nicely for a chain but having a four
coordinate system really complicates things! I have run a very simple
pdb file using pdb2gmx, the new .rtp file above and a handwritten .pdb
file with 2 monomers.
The result is that pdb2gmx is creating extra bonds between the Silicon
of one monomer and the oxygen of the next meaning I am getting a
5-coordinate Silicon.
Pdb2gmx doesn?t seem to be able to distinguish based on bond distances
which oxygens belong to which monomer. The only way I can see past this
is a more elaborate naming system which would introduce yet more
combinations.
So I?m throwing this out as a last resort before I give up. Has anyone
had any success using this method for a similar system? Quartz?
I can see how this rapidly becomes difficult :) I don't believe that pdb2gmx
can handle such "multi-directional" bonding, since the residues that are
connected are not necessarily numerically sequential.
There is, however, another alternative that might work. The specbond.dat file
allows for bizarre connections, provided the atoms are within 10% of a defined
distance cutoff. You probably wouldn't even have to worry about renumbering or
reorganizing your structure.
I would suggest generating a small test case - like a 3x3 grid of tetrahedral
centers, which should encompass all the different types of units (interior,
multiple -OH instead of -O-).
Don't give up, this can surely be accomplished :) Please report back with any
problems or successes. If we can work this out, it would be a fantastic example
for the wiki to help future users.
-Justin
Sorry for my rambling
Jenny
--
========================================
Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin
========================================
_______________________________________________
gmx-users mailing list gmx-users@gromacs.org
http://lists.gromacs.org/mailman/listinfo/gmx-users
Please search the archive at http://www.gromacs.org/search before posting!
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-requ...@gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php