VMD-L Mailing List

From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Sun Nov 15 2009 - 08:46:47 CST

Hi Francesco,
this sounds like it may be the same bug with chain splitting that Luis
Agullo encountered last week. Could you try following the solution at
http://www.ks.uiuc.edu/Research/vmd/mailing_list/vmd-l/14725.html

You need to download the referenced tcl file and then, in a new vmd
session, type
source autopsf.tcl
(pointing to that version of autopsf) prior to using autopsf or cgtools.

Please let me know if you still have trouble.

Best,
Peter

Francesco Pietra wrote:
> 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
>