[PyMOL] Segmentation Fault / Bus Error
Hey Folks - I am trying to assign colors via APBS plugin to a surface which fails with a Bus error on Mac and Segfault on Linux. Versions are 1.1r2pre on Linux and Mac, as well as the latest from svn- trunk on Mac as well... The command that actually fails is ramp_new, as I found out by doing it by hand. Any Ideas where I could look for? Best Martin
[PyMOL] select by absolute coordinate
I quickly checked the selection algebra and property selector wiki pages and couldn't find anything related to this. Is it possible e.g. to select all atoms with (absolute) z - coordinates greater than 5A? Best Martin
[PyMOL] running script on startup
Hey folks, I am trying to fire up a script with pymol -r doit.pml PyMOLrun doit.pml,main Traceback (most recent call last): File /sw/lib/pymol-py25/modules/pymol/parser.py, line 334, in parse parsing .run_file(exp_path(layer.args[0]),__main__.__dict__,__main__.__dict__) File /sw/lib/pymol-py25/modules/pymol/parsing.py, line 455, in run_file execfile(file,global_ns,local_ns) File doit.pml, line 2 select resname SOL ^ SyntaxError: invalid syntax Can't I use these commands or any ideas what I am doing wrong there? :-) Best Martin
Re: [PyMOL] running script on startup
Am 20.02.2009 um 17:54 schrieb Pete Meyer: I am trying to fire up a script with pymol -r doit.pml pymol doit.pml should be sufficient (no need for -r). Oh ok, without -r it works :-) Best Martin
Re: [PyMOL] System freezes in cause of memory abuse
Am Sonntag, 12. Oktober 2008 schrieb Christian Seifert: Hi! I'm working with big structures and even bigger trajectories. When I try to load a trajectory, pymol takes the full 2GB RAM and freezes my PC (ubuntu 7.10/gnome/core 2 duo/2 GB RAM). The only way out is pulling the plug. :-( My Question: Is there a possibility to reduce the memory used by pymol? The nice command did not work, because it just reduces the available CPU time for pymol and not the available memory. On linux systems, you could try ulimit to limit the maximum available memory for current shell and all spawned processes. ulimit -a shows current setup ulimit -v SizeInKB allows setting of a limit in current shell. Best Martin -- Tip of the Day: Nur Vollidioten tun Vollquoten! (Markus Schaber in de.rec.sf.babylon5)
Re: [PyMOL] protein contact potential
Am Freitag, 5. September 2008 schrieb cortines: would anybody know by which calculations pymol generates the proteins' electrostatic maps? Is there a place where I can read about it? I tried the manual and the wikipymol and could not find it. To my knowledge, this can be (only?) done via the APBS plugin which uses the Adaptive Poisson Boltzmann Solver: http://apbs.sourceforge.net/ Best Martin
[PyMOL] Cluster / Cave / Holobench version of PyMol
Hello all, has anybody of you knowledge of efforts to use PyMol on CAVE like setups for 3D visualization? We have a Holobench with a Headtracking System here and I wonder if anybody knows more than I do. The only thing I found is: http://www.gruchalla.org/papers/C4_111_Gruchalla.pdf where CaveLib has been used. Best wishes Martin
[PyMOL] Keeping PDB ordering
Hi all, is there a simple way to prevent reordering of atoms and insertion of additional TER records by pymol? I have a protein embedded in a bilayer with tip4p water, where this occurs when I delete parts and export the rest of the System. Best wishes Martin
Re: [PyMOL] algorithm for alignment
Am Samstag, 26. Januar 2008 schrieb Lu Lin: Hi, Abhi, You mean save them separately? Is there some way to save them into one file? Thanks!! By default, the save command should save all (visible?) atoms. Should be in documentation of save. Any way you can also make a selection of your two molecules and save the selection to one file. Even if you have two files, cat will produce a single file quicker than loading them into pymol ;-) Best Martin
Re: [PyMOL] best linux distribution to run pymol
Am Freitag, 13. Juli 2007 schrieb Joris Beld: Gruezi Boris, I do not want to start a linux distribution war but we are curious under which distribution pymol is developed/tested and what people generally are using to run pymol (with 3D capability!)?? Thanks a lot for the info. Every recent distribution should give you good 3D performance if drivers are set up correctly. I can share OpenSUSE packages (slightly older) than the officially available deb's for ubuntu (in gutsy). They should also work/be available for Debian as well. Best Martin -- Falls dir die Antwort zu ungenau erscheint, könnte es an der Fragestellung liegen. (Daniel Fass in de.org.ccc)
[PyMOL] spectral coloring
Hi there, i am trying to map b factors on a surface representation. I am using spectrum b,blue_white_red,mymol and the colors are fine. Is there a way to fine tune the color spectrum, such as with the Midpoint and Offset sliders in vmd? Best Martin
Re: [PyMOL] coloring based on sequence identity
Am Dienstag, 26. Juni 2007 schrieb Eva Vanamee: Hi, I'd like to color a molecule based on its sequence similarity to another molecule. Can this be done in Pymol? If you can quantify similarity in a number you can write it into the b-factors of the atoms and then color it by them. Best Martin -- Wenn Theorie und Realität nicht übereinstimmen, muß entweder die Theorie oder die Realität geändert werden. Das erste nennt man Wissenschaft, das zweite Religion. [Achim Stoesser, 12.8.1999]
[PyMOL] dimension lock feature
Hi there, just asking if anyone here would need it too? When using the drag tool to move/rotate tool, what about some buttons to constrain movement to specific dimensions. e.g. only rotate arond z-axis, move in x-y-plane and so on. Secondly, a rotation in fixed segments of x degrees, or movement in 1A steps would be great too. What do you think? Or did i miss an already available feature. Cheers Martin
[PyMOL] mutagenesis script
Hi there, does anybody of you has an idea, if the mutagenesis stuff is available for use inside scripts too? I wanna load a pdb, mutate a resid and save the resulting pdb via commandline. The only thing I found was modifying atoms, when I searched via help and also checked the wiki, i hope that I didn't miss sth. Cheers Martin
Re: [PyMOL] PyMOL compilation from source
Am Samstag, 28. April 2007 schrieb Michael Kluge: Would that be feasable? Oh sorry, i overlooked the ia64 in your gcc line. I have no idea how much IA64 SuSE, differs from the others. OpenSuSE is only provided for i586 and x86_64, not Itanium. But if you want, you can either use my spec file and compile yourself or I can also ask the guys here at the LRZ if they can check this. I am using opensuse build service to build for Opensuse 10.2 http://build.opensuse.org/project/monitor?project=home%3Amartoss You could try the IA32 stuff in the Itanium. But that runs through an IA32 emulation layer. Thats like running with ball and chains. For a GUI The emulation layer is really... ... slow :-). Newer Itaniums don't even have hardware support for this feature but do it entirely in SW. is this probably OK as processor utilization is not the key, but for HPC it is just slow. We have 2584 Core AMD Athlon PC Farm. Maybe you want to try deimos? Hmm, still did not get the point? Do you wanna run pymol on a cluster and not on a workstation? Most workstations are intel core or opteron/athlon64 based. Cheers Martin -- Echt frustrierend, wenn man so vor seiner Hardware sitzt und den minütlichen Preisverfall verfolgt... (John Gawe in de.comp misc)
Re: [PyMOL] center of mass and distances
Am Freitag, 2. März 2007 schrieb Filippo Marchioni: Hi all! Two questions: Does anyone know how to read the coordinates (x,y,z) of a selected atom in a protein? Or even better Does anyone know how to visualize the center of mass of the protein, reset the axis (x,y,z) using the centroid as origin and then calculate the distance of a resi or an atom from the new origin? I do this by using gromacs tools, g_traj and then creating a new atom with the coordinates in the xvg. -- Da die vier deutschen U-Bahnen nicht am öffentlichen Straßenverkehr teilnehmen, genügen sie selbstverständlich den Regeln der StVO - nämlich in dem Sinne, daß sie ihnen nicht widersprechen. (Jürgen Hoffmann in de.etc.bahn.stadtverkehr)
Re: [PyMOL] Visualizing Molecular Orbitals in PyMol
Am Donnerstag, 15. Februar 2007 schrieb BuzB: I think a way to do this would be to generate a map, like an electron density map, and then generate a mesh from this, contoured at a given value. A good point to start is maybe to look at Learner's APBS plugin. Also there should be some documentation about the ?OpenDX? format apbs is using for potential maps. It has also isosurface display controls implemented. To do this I would like some help; 1. Given that I know the probability amplitude at all points in a given volume of space, how do I generate a map? Will having sign information complicate matters? Sign information? If you're talking about a simple scalar potential, i see no problems arise. The probability amplitude should be always positive or what do you mean? 2. Given a map object, how do I color the mesh depending on the sign of the values in the map? There are 2 things you can do IMHO. Either creating isosurfaces with fixed color or color sth else e.g. the sas surface of a protein with the potential values. Cheers Martin -- The price of freedom is eternal vigilance., Thomas Jefferson
Re: [PyMOL] image metadata plugin
Am Dienstag, 13. Februar 2007 schrieb DeLano Scientific: We have much work to do to make annotations more general and more accessible... Personally, i just store the session with the scene to reget an impression what I did, but you're right this does not explain what you actually see. Cheers Martin -- Liberty is always dangerous, but it is the safest thing we have. (Harry Emerson Fosdick)
[PyMOL] OT: platform discussion
Am Samstag, 20. Januar 2007 00:55 schrieb DeLano Scientific: Hi all, Of course not: cost is a pragmatic consideration, but cost should not solely be defined in terms of purchase price. You must also include costs in terms of your time and frustration. (However, having once been a graduate student, purchase price alone can be THE overriding factor at that time in your life...) I agree with you that cost in terms of money is a minor factor. But frustration can also be greatly due to the closed source character of the software. E.g. IBM ships no native intel mac client for Tivoli Storage Management. Same is the case for linux but here, you can easily set up a 32bit library set or java-vm which satifies the requirements. By ideological reasons, I mean, for example, preferring Linux over MacOS because it is pure free software as per the FSF dogma whereas OS X effectively combines open-source software in synergy with proprietary code. At least for me, noble motives aren't the reason why i prefer linux. I need a flexible system that adapts to my workflow and not the other way round. I am using kile as my favourite tex editor. We tried to install that thing for a colleague of me on his mac. It was pain in the ***. Although macos root's are those of an unix system installation of not precompiled software is not as easy as it sounds. It is a very good thing that some noble individuals are willing to accept less convenience and extra hassles in order to keep Linux alive on the desktop, thus supporting the many positive long-term impacts open-source code and free software have on society. Convenience is a subjective term. Everybody sees it's favorite system/editor/wordprocessor/... as the most convenient one. Habits play also a very big role in considering sth convenient. One thing that is very often difficult to explain to otheres is that many people do not use linux for ideological or monetary reasons. If one thinks back two or three years or so, linux was the only pc-compatible 64Bit os. Or if you need a supercomputer compatible os for your reserarch work, most of them here run on linux and it's nice to have a compatible system for local tests. Mac OS always forces you to use Apple's hardware and not everyone is happy with it. Personally i like the good old ibm-quality of my t40 thinkpad. Others prefer ugly sony stuff etc. so mac os is not a choice for everyone. Several commercial sofware performs bad on mac os due to emulation through rosetta... Everybody has probably good reasons why he uses osX osy and osz. Cheers Martin
[PyMOL] surface coloring fails with gradient loaded
Hi there, i am trying to render fieldlines and sf potential together in a movie. The movie setup is like this: import os set gradient_spacing=5 load ProtOut.pdb,prot load trajshort.xtc,prot mset 1 -7 show surface set surface_solvent=1 for a in range(1,7): cmd.mdo(a,delete flines;delete dxmap;delete dxmap2;delete rmp;delete rmp2;save prot.pdb,prot;os.system(\pdb2pqr --ff=amber prot.pdb prot.pqrlog.txt\);os.system(\apbs apbs.inlog.txt\);load apbsout.dx,dxmap;load apbsout.dx,dxmap2;ramp_new rmp,dxmap, [-1.5,-0.75,0];ramp_new rmp2,dxmap,[-10,0,10];gradient flines,dxmap,-10,10,prot,carve=15;color rmp,flines;set surface_color=rmp,prot) set ray_trace_frames=1 set cache_frames=1 mclear Is there anything wrong on this setup? the surface is always white. If i do a show surface in every frame, then it seems to me as if the surface is calculated for all frames (600 in total). This is too slow to do it on every step. Any ideas? Cheers Martin
[PyMOL] seperate clipping planes for each object
Hi all, is it possible to change clipping planes for each object? Could't find anything related about it. Cheers Martin -- HTML erhöht den Informationsgehalt eines Postings *immer* um ein paar unvorteilhafte Informationen über den Verfasser. (Thore Tams, de.soc.netzkultur, 17.5.1999)
[PyMOL] weird crashes inside /home/$USER
Hi everybody,
i install pymol via svn ebuild (attached). It crashes with a segfault but only
if i am inside my home directory. If i am outside, everything works fine.
Any ideas what this could be related to?
OpenGL graphics engine:
GL_VENDOR: NVIDIA Corporation
GL_RENDERER: GeForce 6600/PCI/SSE2
GL_VERSION: 2.1.0 NVIDIA 96.29
Adapting to GeForce hardware.
Detected 2 CPUs. Enabled multithreaded rendering.
/usr/bin/pymol: line 6: 12230 Segmentation fault python
${PYMOL_PATH}/__init__.py $*
pymol start script says...
mar...@osgiliath ~ $ cat /usr/bin/pymol
#!/bin/sh
PYMOL_PATH=/usr/lib/python2.4/site-packages/pymol
PYMOL_DATA=/usr/share/pymol/data
PYMOL_SCRIPTS=/usr/share/pymol/scripts
export PYMOL_PATH PYMOL_DATA PYMOL_SCRIPTS
python ${PYMOL_PATH}/__init__.py $*
... which looks good to me.
Cheers
Martin
# Copyright 1999-2006 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header:
/var/cvsroot/gentoo-x86/sci-chemistry/pymol/pymol-0.99_rc6-r2.ebuild,v 1.3
2006/10/22 18:41:55 ribosome Exp $
inherit subversion distutils eutils multilib
MY_PV=${PV/_}
MY_S_P=${PN}-${MY_PV}
MY_PV=${MY_PV/./_}
MY_P=${PN}-${MY_PV}
DESCRIPTION=A Python-extensible molecular graphics system.
HOMEPAGE=http://pymol.sourceforge.net/;
#SRC_URI=mirror://sourceforge/pymol/${MY_P}-src.tgz
ESVN_REPO_URI=https://svn.sourceforge.net/svnroot/pymol/trunk/pymol/;
#ESVN_PATCHES=vmdplugin-pymol-svn.diff
LICENSE=PSF-2.2
IUSE=vmdplugins
SLOT=0
KEYWORDS=~amd64 ~ppc ~x86
DEPEND=dev-lang/python
dev-python/pmw
dev-python/numeric
dev-lang/tk
media-libs/libpng
sys-libs/zlib
vmdplugins? ( sci-chemistry/vmd)
media-libs/glut
S=${WORKDIR}/pymol
src_unpack() {
subversion_src_unpack
cd ${S}
# Turn off splash screen. Please do make a project contribution
# if you are able though.
[[ -n $WANT_NOSPLASH ]] epatch ${FILESDIR}/nosplash-gentoo.patch
use vmdplugins epatch ${FILESDIR}/vmdplugin-pymol-svn.diff
# Respect CFLAGS
sed -i \
-e s:\(ext_comp_args=\).*:\1[]:g \
${S}/setup.py
}
src_install() {
python_version
distutils_src_install
cd ${S}
#The following three lines probably do not do their jobs and should be
#changed
PYTHONPATH=${D}/usr/$(get_libdir)/site-packages ${python} setup2.py
# Make our own wrapper
cat ${T}/pymol EOF
#!/bin/sh
PYMOL_PATH=/usr/lib/python${PYVER}/site-packages/pymol
PYMOL_DATA=/usr/share/pymol/data
PYMOL_SCRIPTS=/usr/share/pymol/scripts
export PYMOL_PATH PYMOL_DATA PYMOL_SCRIPTS
${python} \${PYMOL_PATH}/__init__.py \$*
EOF
exeinto /usr/bin
doexe ${T}/pymol
dodoc DEVELOPERS CHANGES
mv examples ${D}/usr/share/doc/${PF}/
dodir /usr/share/pymol
mv test ${D}/usr/share/pymol/
mv data ${D}/usr/share/pymol/
mv scripts ${D}/usr/share/pymol/
}
--- modules/pymol/invocation.py.orig 2003-06-11 13:40:13.0 -0400
+++ modules/pymol/invocation.py 2003-06-11 13:40:29.0 -0400
@@ -179,5 +179,3 @@
else:
options.deferred.append(a)
- if options.show_splash and not options.no_gui:
- options.deferred.insert(0,_do__ cmd.splash(1))
--- /tmp/setup.py 2006-10-30 20:32:59.0 +0100
+++ setup.py 2006-10-30 20:33:31.0 +0100
@@ -69,8 +69,8 @@
/usr/include/freetype2,
# /users/warren/ext/include,
# VMD plugin support
-# contrib/uiuc/plugins/include,
-# contrib/uiuc/plugins/molfile_plugin/src,
+ contrib/uiuc/plugins/include,
+ contrib/uiuc/plugins/molfile_plugin/src,
]
libs=[GL,GLU,glut,png,z,freetype
]
@@ -85,7 +85,7 @@
# Numeric Python support
# (_PYMOL_NUMPY,None),
# VMD plugin support
-# (_PYMOL_VMD_PLUGINS,None),
+ (_PYMOL_VMD_PLUGINS,None),
(_HAVE_LIBPNG,None)]
ext_comp_args=[-ffast-math,-funroll-loops,-O3]
ext_link_args=[]
@@ -230,52 +230,52 @@
layer5/main.c,
# uncomment below for VMD molfile plugin support
# (incomplete at present -- only TRJ, TRR, XTC, DCD so far...)
-# contrib/uiuc/plugins/molfile_plugin/src/PlugIOManagerInit.c,
-# contrib/uiuc/plugins/molfile_plugin/src/avsplugin.cpp,
-# contrib/uiuc/plugins/molfile_plugin/src/bgfplugin.cpp,
-# contrib/uiuc/plugins/molfile_plugin/src/binposplugin.c,
-# contrib/uiuc/plugins/molfile_plugin/src/biomoccaplugin.cpp,
-# contrib/uiuc/plugins/molfile_plugin/src/brixplugin.cpp,
-# contrib/uiuc/plugins/molfile_plugin/src/carplugin.c,
-# contrib/uiuc/plugins/molfile_plugin/src/ccp4plugin.cpp,
-# contrib/uiuc/plugins/molfile_plugin/src/corplugin.c,
-# contrib/uiuc/plugins/molfile_plugin/src/cpmdplugin.c,
-# contrib/uiuc/plugins/molfile_plugin/src/crdplugin.c,
-#
Re: [PyMOL] segmentation fault after switching mainwindow between visualisation and commandline
... i had a similar Problem this morning. What is your PYTHONPATH? For me, unsetting it solved the problems. Cheers Martin
Re: [PyMOL] pymol licensing questions
Am Donnerstag, 9. November 2006 13:13 schrieb bgbg bg: After searching the pymol site for two hours I'm still very confused about the license issue. I understand that using (and from recently - downloading) the binary PyMol package requires subscription. Now what about the source code? May I compile it and use it without any subscription? May I re-distribute the compiled program? Yes, you can checkout the svn code and compile it on your own. AFAIK most of the code is GPL licenced (Warren, correct me if i am wrong). Familiarize yourseslf with the GPL if you're unsure about redistribution of your own code/binaries. There is some thirdparty stuff, which I am not sure about the license. May I make changes to the source code and distribute the derived product? If I may, under which conditions? Yes you can redistribute code and binaries under the conditions of the GPL. Although you need to provide (your modified) code upon request, which means that you have to give back your modified code to the community, unless your're just using it by yourself (in contrast to BSD licences). Although I am not a developer, i would encourage everybody who is using this program as a professional (university, research) to buy a subscription to support work on pymol in future. There's always work that not everybody likes to do, such as documentation and with more money, this task could be assigned to a payed developer. Hope that did answer your questions. Cheers Martin -- Falls dir die Antwort zu ungenau erscheint, könnte es an der Fragestellung liegen. (Daniel Fass in de.org.ccc)
[PyMOL] cones in pymol?
Hello everybody, i found some cone related stuff in pyopengl (in contrib dir of pymol). Is it usable from pymol in some way? It would be great to visualize vector-fields with 3d-Arrows but you need a cone and a cylinder to construct them. Probably the performance when setting up through triangles is or it looks really bad if they're to few triangles. Cheers Martin -- Tipp- und Rechtschreibfehler sind absichtlich eingebaut und dienen der Belustigung des Lesers :-)
Re: [PyMOL] cones in pymol?
Am Mittwoch, 1. November 2006 21:37 schrieben Sie: Hi Zheng, In my PyMOL scripts for our ANM online server, http://ignmtest.ccbb.pitt.edu/cgi-bin/anm/anm1.cgi we are using CGO in PyMOL to show the vector field. I enclose an example in the attachment. You can run the *.py file in PyMOL. looks good, maybe a bit too glossy. The reflections are very strong but this is probably solvable some way. The shape looks very natural Cheers Martin -- Bald werden die Lehrer sogar auf DVD umschwenken müssen, weil nachwachsende Schülergenerationen ihnen den Umgang mit dem Videorecorder nicht mehr erklären können. [Michael Hemmer, de.etc.sprache.deutsch, 22.12.2003]
Re: [PyMOL] Support for gromacs trajectories and cube files
...got the svn code compiled with -D VMD... If i load a PDB, then load a topology and then try to load a trajectory via load loading the trajectory fails, complaining like this: CmdLoad: /home/martin/workspace/testdata/ProtOut.pdb loaded as ProtOut. PyMOLload /home/martin/workspace/testdata/topol.tpr CmdLoad: /home/martin/workspace/testdata/topol.tpr loaded as topol.tpr. PyMOLload /home/martin/workspace/testdata/ProteinTraj.xtc Best wishes, Martin -- Meinungsfreiheit heißt nicht, dass ich mir jeden Scheiß anhören muss. (Tobias Erle in bln.misc)
Re: [PyMOL] PyMOL and Suse10.0
Am Dienstag, 24. Oktober 2006 14:13 schrieb Neil Ranson: So in summary, there is no inherent problem with PyMOL, Suse 10.0 x86_64, nvidia cards and even recent beta drivers to the best of my knowledge. Hi Neil, what's exactly the output of pymol when it's crashing. Are you using the nvidia or the nv driver? I had a crashing problem today too, because the nvidia driver didn't match the nvidia-glx libs. Best regards, Martin
