[gmx-developers] pdb2gmx patch problems

[Roland Schulz]

Hi,

I have three problems with pdb2gmx. They are all related to the n-terminus
patch I use for cellulose.

The patch is:
[ replace ]
C4              CTS     12.01100        0.14
H4              HAS     1.00800         0.09
[ add ]
1     2         O4      C4      C5      C6
      OHS    15.99940   -0.660
1     2         HO4     O4      C4      C5
      HOS     1.00800   0.430
[ bonds ]
O4      C4

The problems are:

1)

I add two atoms in the N Terminus Patch like this:

Still pdb2gmx compalns:
Fatal error:
Atom O4 in residue BGLC 1 not found in rtp entry with 21 atoms
             while sorting atoms

It works without problem with -nosort option. The problem is that the
function sort_pdbatoms is called by pdb2gmx before the patches are applied.
IMHO the reason this is no problem for a protein is that the additional
atoms are renamed using xlateat.dat so that they are known even before the
patches are applied.


2)
The charge and charge group number of two hydrogen atoms get changed even
though they are not part of the patch at all.

The residue is:
[ BGLC ]
 [ atoms ]
        C1      CBS     0.3     1
        H1      HAS     0.1     1
        O1      OES     -0.4    1
        C5      CTS     0.1     1
        H5      HAS     0.1     1
        O5      OES     -0.4    1
        C4      CTS     0.1     1
        H4      HAS     0.1     1
        C2      CTS     0.140   2
        *H2      HAS     0.090   2*
        O2      OHS     -0.660  2
        HO2     HOS     0.430   2
        C3      CTS     0.140   3
       * H3      HAS     0.090   3*
        O3      OHS     -0.660  3
        HO3     HOS     0.430   3
        C6      CPS     0.050   4
        H61     HAS     0.090   4
        H62     HAS     0.090   4
        O6      OHS     -0.660  4
        HO6     HOS     0.430   4

but after the patch the charge and charge group numbers of H2 and H3  are:
     1        CTS      1   BGLC     C5      1        0.1     12.011   ; qtot
0.1
     2        HAS      1   BGLC     H5      1        0.1      1.008   ; qtot
0.2
     3        OES      1   BGLC     O5      1       -0.4    15.9994   ; qtot
-0.2
     4        HAS      1   BGLC     H1      1        0.1      1.008   ; qtot
-0.1
     5        CBS      1   BGLC     C1      1        0.3     12.011   ; qtot
0.2
     6        OES      1   BGLC     O1      1       -0.4    15.9994   ; qtot
-0.2
     7        CTS      1   BGLC     C2      2       0.14     12.011   ; qtot
-0.06
     8        HAS      1   BGLC    * H2      3        0.1*      1.008   ;
qtot 0.04
     9        OHS      1   BGLC     O2      4      -0.66    15.9994   ; qtot
-0.62
    10        HOS      1   BGLC    HO2      4       0.43      1.008   ; qtot
-0.19
    11        CTS      1   BGLC     C3      5       0.14     12.011   ; qtot
-0.05
    12        HAS      1   BGLC     *H3      6        0.1*      1.008   ;
qtot 0.05
    13        OHS      1   BGLC     O3      7      -0.66    15.9994   ; qtot
-0.61
    14        HOS      1   BGLC    HO3      7       0.43      1.008   ; qtot
-0.18
    15        CTS      1   BGLC     C4      8       0.14     12.011   ; qtot
-0.04
    16        HAS      1   BGLC     H4      8       0.09      1.008   ; qtot
0.05
    17        OHS      1   BGLC     O4      8      -0.66    15.9994   ; qtot
-0.61
    18        HOS      1   BGLC    HO4      8       0.43      1.008   ; qtot
-0.18
    19        CPS      1   BGLC     C6      9       0.05     12.011   ; qtot
-0.13
    20        HAS      1   BGLC    H61      9       0.09      1.008   ; qtot
-0.04
    21        HAS      1   BGLC    H62      9       0.09      1.008   ; qtot
0.05
    22        OHS      1   BGLC     O6      9      -0.66    15.9994   ; qtot
-0.61
    23        HOS      1   BGLC    HO6      9       0.43      1.008   ; qtot
-0.18

The charges and charge groups are correct for all but the first residue. So
for all residues for which the n-terminus patch is not applied.

3)
The above add of two atoms misses one bond. I need to specify manually to
add the bond (as shown above):
[ bonds ]
O4      C4


Non of the problems is very severe, because I have a workaround for 1 and 3
and I can manually change 2. But I wanted to report the problems.
Roland


-- 
ORNL/UT Center for Molecular Biophysics cmb.ornl.gov
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-developers/attachments/20090126/5b734fe0/attachment.html>