Hi Jeremy,
If the proteins have significant homology, then you can use the align command:
align prot1ca,prot2
which will perform a sequence alignment of prot1 against prot2, and then an
optimizing fit using the CA positions. I'm not sure if the help text for align
got into 0.82, but the next version will definitely have it.
PyMOL's command length is limit to about 1000 characters, which is probably
what is causing pair_fit to crash. However, ideally it should complain
gracefully instead of crashing, so if you have a sample script input PDB
files which cause it to crash, please send them to me as attachments.
As far as script options go, you've pointed out a hole which I had forgotten
about, but Robert's suggestion on defining a function call in the script is
definitely the way to go for now.
Cheers,
Warren
--
mailto:war...@sunesis.com
Warren L. DeLano, Ph.D.
Informatics Manager
Sunesis Pharmaceuticals, Inc.
341 Oyster Point Blvd.
S. San Francisco, CA 94080
(650)-266-3606 FAX:(650)-266-3501
-Original Message-
From: Jeremy Craven [mailto:c.j.cra...@sheffield.ac.uk]
Sent: Thursday, September 26, 2002 2:00 AM
To: pymol-users@lists.sourceforge.net
Subject: [PyMOL] pair_fit crashes
I have been trying to write a script to overlay two arbitrary proteins via my
own alignment and the 'pair_fit' command.
1) After running into the problem I described below I looked through the
pymol-users mailing list as hard as I could and found a statment that such
arbitrary alignment should be in pymol anyhow from vsn 0.80. I am using 0.82
but cannot find it. Is it there ?
2) Anyhow, (and assuming that I might want to do alignments my own way anyhow),
I created a script which read in an alignment and created a big long pair_fit
command of the form
pair_fit (tbr and resi 187 and name ca),(bs and resi 170 and name ca), (tbr and
resi 192 and name ca),(bs and resi 175 and name ca), etc
The script then executed this string with cmd.do
If I restricted this to about 10 atom pairs it worked fine. If I allow more
atoms it crashes the program. On linux it gave segmentation violation. On
irix it gave bus error. On irix it gave another message that implied that
maybe the selections had got truncated. On linux it got as far as giving a
sensible 'ExecutiveRMS' message (including a value for the rmsd and the number
of atoms) and then froze and eventually core dumped. It therefore seems that
the long selection has been accepted properly, but may it corrupts something
else as a side effect ?
Any ideas anyone ?? I wondered about doing all the selections invidually into
named selections and then create a shorter pair_fit command, but decided I
would ask here first.
A second little quesion is how do you send command line arguments to a pymol
python script.
If I say
PyMOL run test1.py abcdef
it says
Traceback (most recent call last):
File /usr/lib/python1.5/site-packages/pymol/modules/pymol/parser.py, line
186, in parse
execfile(args[nest][0],pymol_names,pymol_names)
IOError: [Errno 2] No such file or directory
whereas test1.py really does exist, and run test1.py works fine.
Otherwise pymol is a great joy to have discovered in the last month or so, and
a project that I hope I can support (e.g. by contributing scripts like this one
once I can get it to work properly ...)
Cheers
Jeremy
--
*
Dr C. Jeremy Craven
Department of Molecular Biology and Biotechnology
University of Sheffield,
Firth Court, Western Bank
S10 2TN Sheffield UK
e-mail: c.j.cra...@shef.ac.uk
http://www.shef.ac.uk/uni/projects/nmr/CJC/CJC.html
Phone: x24323
From outside Sheffield: 0114 222 4323
From outside UK: +44 114 2224323
Fax: 0114 272 8697
*
