[gmx-users] Sugar topologies (complete message)
Lieven Buts
lieven at ultr.vub.ac.be
Thu Jan 31 10:31:23 CET 2002
(Apologies for posting an incomplete version of this message
earlier.)
Hello all,
I am trying to use Gromacs 3.0 to study the confromational
properties of sugars in solution and bound to proteins. Initially
I wanted to make topology building blocks (analogous to to the
amino acid building blocks) but now I think it might be
simpler to use PRODRG to generate topologies for the sugars I
need.
As a test case I used the mannose(alpha1,4)mannose(alpha1-)methyl
disaccharide:
----- m4m.pdb -------------------------------------------------------
ATOM 3714 C1 MAN S 1 4.529 42.024 25.677 1.00 27.07
ATOM 3715 C2 MAN S 1 3.472 41.123 25.039 1.00 27.92
ATOM 3716 O2 MAN S 1 4.061 40.375 23.988 1.00 29.21
ATOM 3717 C3 MAN S 1 2.906 40.171 26.094 1.00 26.40
ATOM 3718 O3 MAN S 1 1.970 39.287 25.497 1.00 27.87
ATOM 3719 C4 MAN S 1 4.035 39.372 26.738 1.00 21.18
ATOM 3720 O4 MAN S 1 3.505 38.578 27.784 1.00 23.26
ATOM 3721 C5 MAN S 1 5.094 40.327 27.303 1.00 25.69
ATOM 3722 O5 MAN S 1 5.555 41.228 26.268 1.00 24.50
ATOM 3723 C6 MAN S 1 6.311 39.595 27.841 1.00 22.90
ATOM 3724 O6 MAN S 1 7.080 39.028 26.790 1.00 20.69
ATOM 3725 C1 MAN S 2 4.412 46.882 27.747 1.00 47.19
ATOM 3726 O1 MAN S 2 3.168 46.841 28.357 1.00 51.01
ATOM 3727 C2 MAN S 2 4.251 46.618 26.248 1.00 45.91
ATOM 3728 O2 MAN S 2 5.515 46.713 25.609 1.00 49.07
ATOM 3729 C3 MAN S 2 3.669 45.220 26.031 1.00 44.04
ATOM 3730 O3 MAN S 2 3.604 44.936 24.641 1.00 49.15
ATOM 3731 C4 MAN S 2 4.536 44.175 26.730 1.00 39.76
ATOM 3732 O4 MAN S 2 3.920 42.881 26.599 1.00 32.21
ATOM 3733 C5 MAN S 2 4.683 44.537 28.211 1.00 42.72
ATOM 3734 O5 MAN S 2 5.239 45.868 28.345 1.00 44.76
ATOM 3735 C6 MAN S 2 5.596 43.581 28.963 1.00 42.20
ATOM 3736 O6 MAN S 2 6.925 44.075 29.036 1.00 43.13
ATOM 3737 C7 AME S 3 2.916 47.854 29.323 1.00 53.95
END
---------------------------------------------------------------------
PRODRG returned a Gromacs topology file. I modified the residue and
atom names to something a little more meaningful (for me):
----- m4mh.top-------------------------------------------------------
[ defaults ]
; nbfunc comb-rule gen-pairs fudgeLJ fudgeQQ
1 1 no 1.0 1.0
#include "ffgmx.itp"
#include "spc.itp"
[ moleculetype ]
;name nrexcl
M4MM 3
...
[ atoms ]
; nr type resnr resid atom cgnr charge
1 HO 1 M4MM H2A 1 0.000
2 OA 1 M4MM O2A 2 0.000
3 CH1 1 M4MM C2A 3 0.000
4 CH1 1 M4MM C3A 4 0.150
5 OA 1 M4MM O3A 4 -0.548
6 HO 1 M4MM H3A 4 0.398
7 CH1 1 M4MM C4A 5 0.150
8 OA 1 M4MM O4A 5 -0.548
9 HO 1 M4MM H4A 5 0.398
10 CS1 1 M4MM C5A 6 0.000
11 CH2 1 M4MM C6A 7 0.150
12 OA 1 M4MM O6A 7 -0.548
13 HO 1 M4MM H6A 7 0.398
14 OS 1 M4MM O5A 8 0.000
15 CS1 1 M4MM C1A 9 0.180
16 OS 1 M4MM OG 9 -0.360
17 CS1 1 M4MM C4B 9 0.180
18 CH1 1 M4MM C3B 10 0.150
19 OA 1 M4MM O3B 10 -0.548
20 HO 1 M4MM H3B 10 0.398
21 CH1 1 M4MM C2B 11 0.150
22 OA 1 M4MM O2B 11 -0.548
23 HO 1 M4MM H2B 11 0.398
24 CS1 1 M4MM C1B 12 0.180
25 OS 1 M4MM O1B 12 -0.360
26 CH3 1 M4MM C7 12 0.180
27 OS 1 M4MM O5B 13 0.000
28 CS1 1 M4MM C5B 14 0.000
29 CH2 1 M4MM C6B 15 0.150
30 OA 1 M4MM O6B 15 -0.548
31 HO 1 M4MM H6B 15 0.398
[ bonds ]
;ai aj fu c0 c1
1 2 1 0.100 313800.0 0.100 313800.0 ; HAA OAC
...
30 31 1 0.100 313800.0 0.100 313800.0 ; OAW HAG
[ pairs ]
;ai aj fu c0 c1
1 4 1 ; HAA OAC
...
28 31 1 ; OAM CAX
[ angles ]
;ai aj ak fu c0 c1
1 2 3 1 109.5 397.5 109.5 397.5 ; HAA OAC CAB
...
29 30 31 1 109.5 397.5 109.5 397.5 ; CAV OAW HAG
[ dihedrals ]
;ai aj ak al fu c0 c1 m c0 c1 m
3 2 15 4 2 35.3 836.8 0 35.3 836.8 0 ; IDI CAB OAC
CAA CAD
...
28 29 30 31 1 0.0 1.3 3 0.0 1.3 3 ; DI CAT CAV
OAW HAG
[ system ]
ManA4Man
[ molecules ]
M4MM 1
---------------------------------------------------------------------------
I removed the "SOL 0" line, since genbox just adds an additional
"SOL <n>" line at the end.
This topology corresponds to the following (atom types given
below the atom names):
H6A-O6A-C6A H6B-O6B-C6B
HO OA CH2 HO OA CH2
| |
C5A-O5A C5B--O5B
CS1 OS CS1 OS
/ \ / \
H4A-O4A-C4A C1A-OG--C4B C1B--O1B--C7
HO OA CH1 CS1 OS CH1 CS1 OS CH3
\ / \ /
C3A-C2A C3B-C2B
CH1 CH1 CH1 CH1
| | | |
O3A O2A O3B O2B
OA OA OA OA
| | | |
H3A H2A H3B H2B
HO HO HO HO
I compared this to the topology for glucose (an
isomer of mannose) in ${gromacs}/share/top/ffgmx.rtp :
H6--O6--C6
HO OA CS2
|
C5---O5
CS1 OS
/ \
O4-C4 C1
OS CS1 CS1
\ /
C3----C2
CS1 CS1
| |
O3 O2
OA OA
| |
H3 H2
HO HO
This is probably intended as a building block for
a 1-4 glucose polymer. The biggest difference is that
the C2, C3 and C4 atoms are of type CS1 instead of CH1
as in the PRODRG topology, and the C6 atoms are CS2
instead of CH2. If I understand correctly, the CS1 and
CS2 atom types were introduced to better model specific
sugar effects such as the exo-anomeric effect.
Would it be necessary to replace the CH types by
the corresponding CS types?
--
Lieven Buts
Vrije Universiteit Brussel
More information about the gromacs.org_gmx-users
mailing list