Hi Peter
You have uncovered a bug in Chainsaw. If you have built from source the
patch below should fix it. If you installed form a binary, let me know
what operating system you are using and I will try to generate a new
binary for you.
Norman
Index: chainsaw.cpp
===================================================================
RCS file: //ccpdisk/xtal/CVSROOT/ccp4/src/mmdb_app_/chainsaw.cpp,v
retrieving revision 1.27.4.7
diff -c -r1.27.4.7 chainsaw.cpp
*** chainsaw.cpp 9 Apr 2009 08:54:07 -0000 1.27.4.7
--- chainsaw.cpp 12 Aug 2009 10:54:11 -0000
***************
*** 801,806 ****
--- 801,807 ----
if (!strcmp(name2, " CB ")) deleteB = true;
if (!strcmp(name2, " SG ") || !strncmp(name2, " OG", 3) ||
!strncmp(name2, " CG", 3)) deleteG = true;
}
+ residue->TrimAtomTable();
// Set Gamma atom to be the correct type
// Deleting and adding back in ensures any conformation label is
removed
________________________________
From: CCP4 bulletin board [mailto:ccp...@jiscmail.ac.uk] On Behalf Of
Peter Grey
Sent: 11 August 2009 14:01
To: CCP4BB@JISCMAIL.AC.UK
Subject: [ccp4bb] Chainsaw question
Hi all,
I attempt to create a suitable model for molecular replacement using
CHAINSAW.
Whenever the program needs to mutate a Proline into something other than
alanine I get "segmentation fault".
This happens if the order of atoms in the proline residue is
(N,CD,CA,CB,CG,C,O).
Could you suggest how to solve this problem or alternatively how to
change all prolines (there are many) to have another order of atoms
which does not pose a problem to chainsaw (N,CA, C,O,CB,CG,CD) ?
Thank you in advance for your advice,
Peter.
--
Scanned by iCritical.
