[MMTK] Bug (?) in reading/writing PDBs

[Urban Hafner]

Hej,
when I run the following script I get an error:

urban at bettong:~/Uni/ws0304/pp/shackle/t$ cat bug.py
import MMTK.PDB
import sys

conf = MMTK.PDB.PDBConfiguration(sys.argv[1])
chains = conf.createPeptideChains()
del conf

out = MMTK.PDB.PDBOutputFile(sys.argv[1]+".new")
out.write(chains)
out.close()

aconf = MMTK.PDB.PDBConfiguration(sys.argv[1]+".new")
achains = aconf.createPeptideChains()
del aconf

The error looks like this:

urban at bettong:~/Uni/ws0304/pp/shackle/t$ python bug.py pdbs/1aru.ent
Traceback (most recent call last):
   File "bug.py", line 13, in ?
     achains = aconf.createPeptideChains()
   File "/sw/lib/python2.3/PDB.py", line 168, in createPeptideChains
     if not line.strip():
   File  
"/sw/src/root-mmtk-2.2-1/sw/lib/python2.3/site-packages/MMTK/ 
Proteins.py", line 243, in __init__
   File  
"/sw/src/root-mmtk-2.2-1/sw/lib/python2.3/site-packages/MMTK/ 
Biopolymers.py", line 79, in _setupChain
   File "/sw/lib/python2.3/PDB.py", line 33, in applyTo
     answer = None
   File "/sw/lib/python2.3/PDB.py", line 309, in setResidueConfiguration
     if len(parts) == 1:
IOError: Atom O2 of PDB residue PRO not found in residue Pro of object  
.Pro336
urban at bettong:~/Uni/ws0304/pp/shackle/t$

The PDB file is 1aru from PDB. But this doesn't just happen with this  
file but
with every PDB that ends with Proline. If you compare the two pdbs you  
see this
difference:

Original PDB:

ATOM   2458  N   PRO   344      27.106  10.396  26.216  1.00 12.65       
1ARU2629
ATOM   2459  CA  PRO   344      27.375   9.290  27.115  1.00 13.54       
1ARU2630
ATOM   2460  C   PRO   344      26.774   9.343  28.505  1.00 14.51       
1ARU2631
ATOM   2461  O   PRO   344      26.511  10.416  29.029  1.00 14.14       
1ARU2632
ATOM   2462  CB  PRO   344      28.878   9.208  27.125  1.00 14.06       
1ARU2633
ATOM   2463  CG  PRO   344      29.276  10.641  26.990  1.00 15.25       
1ARU2634
ATOM   2464  CD  PRO   344      28.309  11.168  25.947  1.00 12.48       
1ARU2635
ATOM   2465  OXT PRO   344      26.583   8.288  29.106  1.00 17.97       
1ARU2636
TER    2466      PRO   344

In the file that MMTK (or is it Scientific Python?):

ATOM   4839  C   PRO A 336      26.774   9.343  28.505  0.00  0.00       
      C
ATOM   4840 2HB  PRO A 336      29.246   8.785  28.060  0.00  0.00       
      H
ATOM   4841 2HD  PRO A 336      28.663  10.981  24.933  0.00  0.00       
      H
ATOM   4842  CB  PRO A 336      28.878   9.208  27.125  0.00  0.00       
      C
ATOM   4843  CA  PRO A 336      27.375   9.290  27.115  0.00  0.00       
      C
ATOM   4844  CG  PRO A 336      29.276  10.641  26.990  0.00  0.00       
      C
ATOM   4845 3HB  PRO A 336      29.249   8.617  26.287  0.00  0.00       
      H
ATOM   4846  N   PRO A 336      27.106  10.396  26.216  0.00  0.00       
      N
ATOM   4847  CD  PRO A 336      28.309  11.168  25.947  0.00  0.00       
      C
ATOM   4848 3HD  PRO A 336      28.115  12.233  26.074  0.00  0.00       
      H
ATOM   4849 2HG  PRO A 336      30.309  10.725  26.652  0.00  0.00       
      H
ATOM   4850 3HG  PRO A 336      29.161  11.165  27.939  0.00  0.00       
      H
ATOM   4851  HA  PRO A 336      26.873   8.395  26.748  0.00  0.00       
      H
ATOM   4852  O2  PRO A 336      26.583   8.288  29.106  0.00  0.00       
      O
ATOM   4853  O1  PRO A 336      26.511  10.416  29.029  0.00  0.00       
      O
TER    4854      PRO A 336
END

The difference seems to be between OXT in the original file and O1 and  
O2
in the second one.

So is this a bug in MMTK or in Scientific Python and is there something  
I
can do about it?

Urban

PS: I'm using Scientific Python 2.4 and MMTK 2.2