From: Dong Luo (
Date: Wed Dec 22 2010 - 09:51:39 CST

autopsf may have a problem recognize the CT2 format in the pdb file. Try create
a new pdb file by removing the CT2 lines and let autopsf create the CT2 by using
the patch "last CT2". Then check the correct format and modify your original pdb
file accordingly.


----- Original Message ----
From: Francesco Pietra <>
Sent: Wed, December 22, 2010 1:50:35 AM
Subject: Fwd: vmd-l: autopsf error: failed on end of segment

To be more specific, autopsf recognizes the CT2 segment as
non-protein. Therefore, in the gui I edit:

-- the PHE segment 1-81, adding 82-86 in a single segment 1-86

--N terminal patch, from NTEr to CT2

--C terminal patch from CTER to CT2

-- delete the 82-86 segment

--Selection from 'everything' to 'protein'

The display now shows the entire PHE-CT2

I have tried other combinations in the gui but always ending in "
error: failed on end of segment". Either (very likely) my procedure is
incorrect or autopsf gets confused by CT2. I have considered related
problems on the VMD mail list.



---------- Forwarded message ----------
From: Francesco Pietra <>
Date: Tue, Dec 21, 2010 at 6:49 PM
Subject: vmd-l: autopsf error: failed on end of segment

With VMR 1.8.7 I am trying to generate psf for a protein ending with
amidated phenylalanine, using validated top_all27_prot_na.rtf. I tried
also 'source autopsf.tcl', using a recent script, before autopsf. It
always crashes with error: failed on end segment.

The pdf file ends with (here I tried separately chain D, renumbering
atoms) using the method for amber:

ATOM 80 HE2 PHE D 4 -31.197 22.536 3.559 1.00 0.00 4SG H
ATOM 81 HZ PHE D 4 -30.544 24.713 4.555 1.00 0.00 4SG H
ATOM 82 C CT2 D 5 -32.820 18.877 8.926 1.00 0.00 4SG C
ATOM 83 O CT2 D 5 -34.065 18.451 9.222 1.00 0.00 4SG O
ATOM 84 NT CT2 D 5 -31.788 18.180 9.397 1.00 0.00 4SG N
ATOM 85 HT2 CT2 D 5 -30.844 18.477 9.196 1.00 0.00 4SG H
ATOM 86 HT1 CT2 D 5 -31.949 17.354 9.956 1.00 0.00 4SG H

Is that correct?

Thanks for help

francesco pietra