Error: failed on end of segment with RBCG

From: Francesco Pietra (chiendarret_at_gmail.com)
Date: Sun Nov 15 2009 - 03:18:57 CST

I am unable to carry out a RBCG with a regular pdb file by using the
Cgtool plugin 1.0 in vmd 1.8.7. Always Error: failed on end of
segment; molecule destroyed. While the intermediate atoms are
correctly identified, the final atom is not and correcting it for rbcg
did not help.

Details. The input pdb file is made of three subunits labeled A, B, C.
The intermediate lines between subunits, and end of file read:

ATOM   3421  O   LYS A 426     150.365  56.883  75.135  0.00  0.00           O
ATOM   3422  OXT LYS A 426     152.235  56.235  74.089  0.00  0.00           O
TER    3423      LYS A 426
ATOM   3424  N   LEU B   1     148.151  91.377  78.994  0.00  0.00           N

ATOM   6825  O   LYS B 424     153.891  87.609  92.504  0.00  0.00           O
ATOM   6826  OXT LYS B 424     151.904  88.373  93.007  0.00  0.00           O
TER    6827      LYS B 424
ATOM   6828  N   LEU C   1     141.730  74.503 111.620  0.00  0.00           N

ATOM  10196  O   TYR C 420     150.722  63.772 103.158  0.00  0.00           O
ATOM  10197  OXT TYR C 420     149.062  65.176 103.644  0.00  0.00           O
END

Opening this pdb file in vmd, Extension... Modeling...AutoPSF, using
topology file autoprovided ../plugins/noarch/tcl/readcharmtop1.0, the
segments (subunits) are identified as follows

P1 426 1- 3422 NTER CTER Prot
P2 424 3423- 6825 NTER CTER Prot
P3 420 6826- 17021 NTER CTER Prot

i.e., ATOM 10197 is incorrectly identified. Correcting 17021 to 10197
in the AUTOPSF window, and hitting "Create chains", the message
appears: "Because no patches were applied, a complete psf and pdb will
be generated". Hitting OK, both autopsf and autopdb are generated. In
comparison to the above input pdb file, now the autopdb reads:

ATOM 6763 HZ3 LYS A 426 149.113 57.925 74.526 0.00 0.00 P1 H
ATOM 6764 N LEU B 1 148.151 91.377 78.994 1.00 0.00 P2 N

ATOM 13487 HZ3 LYS B 424 153.026 85.538 92.999 0.00 0.00 P2 H
ATOM 13488 N LEU C 1 141.730 74.503 111.620 1.00 0.00 P3 N

ATOM 20139 HE2 TYR C 420 146.279 58.147 104.781 0.00 0.00 P3 H
END

I don't think this is useful for coarse graining but it shows that
correting the ATOM guess it worked for psf generation. It does not for
rbcg generation, starting from either the generated autopdb file or
the original pdf file.

First, starting from autopdb/autopsf. Loading
./cgtools1.0/protein.cgc, the cgpdb file is generated although no psf
file had been provided. The cgpdf file, comparatively to the above
readings, reads:

ATOM 826 LYS2 LYS A 426 149.111 59.319 73.405 1.00 0.00 P1 C
ATOM 827 BB LEU B 1 147.436 90.187 79.169 1.00 0.00 P2 C

ATOM 1648 LYS2 LYS B 424 153.292 86.731 95.010 1.00 0.00 P2 C
ATOM 1649 BB LEU C 1 142.032 75.308 110.517 1.00 0.00 P3 C

ATOM 2461 BB TYR C 420 149.061 63.779 103.848 1.00 0.00 P3 C
ATOM 2462 TYR2 TYR C 420 146.102 60.027 103.195 1.00 0.00 P3 C
END

Opening this cgpdf in vmd ... AUTOPSF .. topology rbcg-2007.top, guess
and split chains
 reads

P1 426 1- 826 NTER CTER Prot
P2 424 827- 1648 NTER CTER Prot
P3 420 1649- 4111 NTER CTER Prot

where ATOM 2462 is incorrectly guessed Correcting 4111 to 2462,
hitting "Create chins" gave "Error: failed on end of segment".
-----------
Second, starting from the original pdb file, opening in vmd,
Extensin.. Modeling...CG Builder.. RB, either bloading protein.cgc or
rb-2007.top, Error: atomselect: cannot parse selection text.

I wonder whether something has to be loaded to deal with atom OXT, or
the procedure I followed is wrong.

Thanks for help

francesco pietra

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:53:30 CST