AMBER Archive (2007)

Subject: AMBER: ptraj error

From: FyD (fyd_at_q4md-forcefieldtools.org)
Date: Fri Jul 20 2007 - 16:21:17 CDT

Dear Amber users,

I use AMBER8/ptraj under linux & IRIX.
Same problem, I get the following error message:

Checking coordinates: ../md1.mdcrd.gz
*** glibc detected *** ptraj: free(): invalid next size (fast):
0x0000000000a3a070 ***
======= Backtrace: =========
[0x672d67]
[0x6763e7]
[0x41425c]
[0x412a7e]
[0x4301d4]
[0x4216ce]
[0x4131b0]
[0x423bd5]
[0x41c6c5]
[0x4003ea]
[0x654140]
[0x4001e9]
======= Memory map: ========
00400000-00714000 r-xp 00000000 03:02 1471924
/usr/local/amber8/exe/ptraj
00914000-0094f000 rwxp 00314000 03:02 1471924
/usr/local/amber8/exe/ptraj
0094f000-00a4b000 rwxp 0094f000 00:00 0
[heap]
2aaaaaaab000-2aaaaab84000 rwxp 2aaaaaaab000 00:00 0
2aaaaab85000-2aaaaabd4000 rwxp 2aaaaab85000 00:00 0
2aaaac000000-2aaaac026000 rwxp 2aaaac000000 00:00 0
2aaaac026000-2aaab0000000 ---p 2aaaac026000 00:00 0
7fff6afaf000-7fff6afc6000 rwxp 7fff6afaf000 00:00 0
[stack]
ffffffffff600000-ffffffffffe00000 ---p 00000000 00:00 0
[vdso]
Abandon

This error has already been reported for Amber9 @
http://structbio.vanderbilt.edu/archives/amber-archive/2007/1604.php
http://structbio.vanderbilt.edu/archives/amber-archive/2007/1623.php

1) Here is my input:

[fyd_at_labo2 ptraj-test]$ more ptraj11.in
trajin ../md1.mdcrd.gz
center :1-178 mass origin
reference Domain12-1A6Z.pdb
rms reference out RMSD.dat :3-176_at_N,CA,C

- md1.mdcrd.gz is a trajectory of a protein domain which contains 2882 atoms
- my reference "Domain12-1A6Z.pdb" contains 2287 atoms (same number of
aminoacids; but presence of a mutation: one amino-acid has simply a
different side chain)
=> In this case, I get the error message reported above...

2) I slightly change the input just above: instead of "reference
Domain12-1A6Z.pdb", I use "reference ../prmcrd.pdb" (this new
reference contains now 2882 atoms, i.e. the initial structure of the
trajectory)
=> No error message

3) I remove 5 side chain atoms from the reference "Domain12-1A6Z.pdb"
and re-run the input in 1)
=> No error message

4) I remove 5 side chain atoms (different residue) from the reference
"Domain12-1A6Z.pdb" and re-run the input in 1)
=> No error message

5) I remove only 1, 2 or 3 side chain atoms from the reference
"Domain12-1A6Z.pdb" and re-run the input in 1)
=> In this case, I get the error message reported above...

6) I remove 4 or 6 side chain atoms from the reference
"Domain12-1A6Z.pdb" and re-run the input in 1)
=> No error message; here, I am lost...

So it looks like the number of atoms of the reference as some
importance in this problem...

Was this bug corrected ?
I saw nothing about this problem @
http://amber.scripps.edu/bugfixes90.html

Thanks, regards, Francois

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu