- Chimera-dev - RBVI Mailing Lists

Re: [chimera-dev] making a python package
by Eric Pettersen 14 Aug '08

14 Aug '08

Hi Bill, Greg is the real expert here, but he's at SIGGRAPH right now so I'll put out what little I know and let Greg correct anything I get wrong when he has time. Yeah, compiling/linking can be a nightmare when you are trying to use libs you have no control over and that may have been compiled with options you didn't anticipate or from modified sources. That's why we cheat and include everything we depend on (which is one reason Chimera is such a huge download). In particular, we include Tk and Tcl libraries. If you look in Chimera's "foreign" source directory, there is a TclTk subdirectory there. So no worries about out-of-date/incompatible versions and all that. We have another layer of control in that we're not (currently) trying to provide a package that other people can easily compile on their own system. So therefore we don't need a configure step. Our makefiles have conditionals based on the OS to get the right flags for linking and compiling and whatnot for the systems that we ourselves compile on. A lot of that makefile logic is concentrated in the "includable" makefiles found in the mk directory, so that individual makefiles only have to include the appropriate mk makefile and then customize a few variables themselves. We try to use "foreign" source with as little patching as possible so that we can easily incorporate updated versions, so many of the foreign packages (e.g. Tcl/Tk) do have their "native" configure step. Nonetheless, the Chimera makefile that initiates the foreign package build may supply arguments to the configure so that the final build uses the right compiler flags, uses the right build/install location, and so forth. I don't know how much help this has been, but feel free to ask more questions and maybe I'll know the answers. :-) --Eric Eric Pettersen UCSF Computer Graphics Lab http://www.cgl.ucsf.edu On Aug 14, 2008, at 11:54 AM, William Baxter wrote: > Hello Eric, > > I'm trying to make an installation package like Chimera's, in which > our > software installs its own Python. Can you explain how your > installation > package is set up? I have a script that goes through the usual > configure, make, and make install commands, but it encounters > different > problems on each platform (for us, Linux and Irix). How does the > Chimera > install make sure everything is linked together properly? > > Any help is greatly appreciated. > > Thanks, > Bill Baxter > > -- > William T. Baxter, Ph.D. > Wadsworth Center > Empire State Plaza, PO Box 509 > Albany, NY 12201-0509 > > > Subject: Re:[chimera-dev] making a python package > From: William Baxter > Date: 02:42 PM > To: Greg Couch > > > Greg, > thanks for your description. I have another couple of questions. > > My script goes through the usual configure, make, make install > commands, > but Chimera seems to go straight to compilation. How do you set it > up to > be so fast? > > Also, finding Tcl/Tk for Tkinter is a headache that seems to be > different on each platform. How do you ensure that Python finds that > right Tcl libs? > > The Chimera installation is really admirable and efficient. I'd like > to > emulate it as far as possible. > > Any advice you can give is greatly appreciated. > > Thanks, > Bill B. > > Greg Couch wrote: >>> On Wed, 30 Jul 2008, William Baxter wrote: >>> >>>>> Hello, >>>>> >>>>> we would like to package our software, Spire >>>>> (www.wadsworth.org/spider_doc/spider/docs/spire/) the way >>>>> Chimera is >>>>> packaged, with its own Python, Tcl/Tk, and other libraries. Do >>>>> you have >>>>> some documentation, or can you point me toward some resources that >>>>> explain how to go about creating such an installation package? >>>>> >>>>> Thanks, >>>>> Bill Baxter >>> >>> We don't have any documentation, but here's a quick description of >>> what we do binary distributions for chimera: >>> >>> On Windows, we use the Inno Setup program, >>> <http://www.jrsoftware.org/isinfo.php>. >>> >>> For Mac OS X, we build a disk image using the hdiutil program. >>> And we >>> modify the name all of the shared libraries with something like: >>> install_name_tool -id @executable_path/$(LIBNAME) >>> >>> On Linux and other Unix systems, we use a modified self-extracting >>> zip >>> archive where an installation program is run after the files have >>> been >>> extracted. >>> >>> In all cases, when chimera starts up, we set up the environment to >>> get >>> our shared libraries (and clear the environment of things that would >>> hurt us, like PYTHONHOME). >>> >>> Hope this helps, >>> >>> Greg Couch >>> UCSF Computer Graphics Lab > > > -- > William T. Baxter, Ph.D. > Wadsworth Center Empire State Plaza, > PO Box 509 Albany, NY 12201-0509

1 0

making a python package
by William Baxter 01 Aug '08

01 Aug '08

Hello, we would like to package our software, Spire (www.wadsworth.org/spider_doc/spider/docs/spire/) the way Chimera is packaged, with its own Python, Tcl/Tk, and other libraries. Do you have some documentation, or can you point me toward some resources that explain how to go about creating such an installation package? Thanks, Bill Baxter -- William T. Baxter, Ph.D. Wadsworth Center Empire State Plaza, PO Box 509 Albany, NY 12201-0509 IMPORTANT NOTICE: This e-mail and any attachments may contain confidential or sensitive information which is, or may be, legally privileged or otherwise protected by law from further disclosure. It is intended only for the addressee. If you received this in error or from someone who was not authorized to send it to you, please do not distribute, copy or use it or any attachments. Please notify the sender immediately by reply e-mail and delete this from your system. Thank you for your cooperation.

2 1

cElementTree broken in latest production release
by Mark Moll 10 Jul '08

10 Jul '08

Minimal example: from xml.etree.cElementTree import XML XML('<fail>') If you execute this code, you get this error: "cannot load dispath table from pyexpat" -- Mark

2 1

problem with solvate
by Jean-Didier Maréchal 19 Jun '08

19 Jun '08

Hi Guys, I tried the solvate module on the alpha build of the 15/06/2008 on windows. And I have the following error: Running sleap command: C:\Archivos de programa\Chimera2\bin\amber10\exe\sleap -f d:\docume~1\jd\config~1\temp\tmpi8eph7\solvate.cmd (solvate) [gtkleap]$ source leaprc.ff03 (solvate) Error: can not find file leaprc.ff03 in all the search path. Failure running sleap Check reply log for details I made a search of the mentioned file and it does not pop up. Cheers JD -----Mensaje original----- De: chimera-users-bounces(a)cgl.ucsf.edu [mailto:chimera-users-bounces@cgl.ucsf.edu] En nombre de chimera-users-request(a)cgl.ucsf.edu Enviado el: miércoles, 18 de junio de 2008 21:00 Para: chimera-users(a)cgl.ucsf.edu Asunto: Chimera-users Digest, Vol 62, Issue 18 Send Chimera-users mailing list submissions to chimera-users(a)cgl.ucsf.edu To subscribe or unsubscribe via the World Wide Web, visit http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users or, via email, send a message with subject or body 'help' to chimera-users-request(a)cgl.ucsf.edu You can reach the person managing the list at chimera-users-owner(a)cgl.ucsf.edu When replying, please edit your Subject line so it is more specific than "Re: Contents of Chimera-users digest..." Today's Topics: 1. Find contour level to enclose a specified volume (Thomas Goddard) 2. Re: Find contour level to enclose a specified volume (Tom Goddard) ---------------------------------------------------------------------- Message: 1 Date: Tue, 17 Jun 2008 15:44:57 -0700 From: Thomas Goddard <goddard(a)cgl.ucsf.edu> Subject: [Chimera-users] Find contour level to enclose a specified volume To: "'Chimera BB'" <chimera-users(a)cgl.ucsf.edu> Message-ID: <48583E69.5090608(a)cgl.ucsf.edu> Content-Type: text/plain; charset="iso-8859-1" Ed Brignole asked about setting a volume contouring level so that the resulting surface encloses a desired volume (cubic Angstroms). Attached is a Python script that does this in two ways. The first way is fast but approximate finding the level such that the number of enclosed grid points times the voxel volume equals the requested volume. Estimating the enclosed volume of the surface this way can give a much different result than actually computing analytically the volume within the surface. The second approach computes surfaces at different contour levels and calculates analytically the volume. It uses bisection to find the correct contour level. This approach can be slow, about 30 seconds for a 256^3 volume (sfv.mrc virus map) with ~2.5e6 surface triangles. Most of that time (80%) is actually computing the enclosed volume. That should be very fast but the code is checking for holes in the surface in a rather slow way. Both methods use the step size (subsampling) set in the volume dialog. Tom

2 1

Re: [chimera-dev] [Chimera-users] displacing atoms following a given vector
by Eric Pettersen 22 May '08

22 May '08

Hi JD, You would displace the atoms of molecule 'm' by the vector (1,2,3) like so: v = chimera.Vector(1,2,3) for a in m.atoms: a.setCoord(a.coord() + v) --Eric Eric Pettersen UCSF Computer Graphics Lab http://www.cgl.ucsf.edu On May 22, 2008, at 1:12 PM, Jean-Didier Maréchal wrote: > Dear All, > > I have been flying through the developer's guide but can't find what I > look for, sorry. > > I have a given molecule and a vector in cartesian coordinates (e.g. a > normal mode but not necessarily) indicating a direction of > displacement > for the different atoms of the molecules. I would like to write a > script > that displaces the atoms following this vector. I am prettysure that > some modules of chimera should allow me to do that, but I can't find > which ones. Could you give me a hand on this please? > > ALl the best > JD > > > El jue, 22-05-2008 a las 12:00 -0700, chimera-users- > request(a)cgl.ucsf.edu > escribió: >> Send Chimera-users mailing list submissions to >> chimera-users(a)cgl.ucsf.edu >> >> To subscribe or unsubscribe via the World Wide Web, visit >> http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users >> or, via email, send a message with subject or body 'help' to >> chimera-users-request(a)cgl.ucsf.edu >> >> You can reach the person managing the list at >> chimera-users-owner(a)cgl.ucsf.edu >> >> When replying, please edit your Subject line so it is more specific >> than "Re: Contents of Chimera-users digest..." >> >> >> Today's Topics: >> >> 1. Stereo projection (Magali Cottevieille) >> 2. Stereo projection (Steven Ludtke) >> 3. Re: Stereo projection (Greg Couch) >> 4. Re: findhbond/5750 (FindHBond failure) (Eric Pettersen) >> >> >> --------------------------------------------------------------------- >> - >> >> Message: 1 >> Date: Wed, 21 May 2008 16:06:07 -0400 >> From: Magali Cottevieille <mc3077(a)columbia.edu> >> Subject: [Chimera-users] Stereo projection >> To: chimera-users(a)cgl.ucsf.edu >> Message-ID: <483480AF.9020604(a)columbia.edu> >> Content-Type: text/plain; charset=ISO-8859-1; format=flowed >> >> Hi, >> >> Is there any way, with two projectors piloted by two graphic >> cards, to >> send from Chimera the stereo left eye view onto one, and the stereo >> right eye view onto the other one ? These two projectors are meant >> to be >> used with filters and special glasses to create a 3D view. >> Thanks for your help! >> >> -- >> Magali Cottevieille, Ph.D. >> Howard Hughes Medical Institute, >> Department of Biochemistry and Molecular Biophysics, >> 630 W 168th St, P&S Black Building 2-221 >> New York, NY 10032 >> >> Ph: (+1) 212-305-9521 >> Fax: (+1) 212-305-9500 >> >> Email: mc3077(a)columbia.edu >> magali.cottevieille(a)gmail.com >> >> >> >> ------------------------------ >> >> Message: 2 >> Date: Wed, 21 May 2008 16:18:26 -0500 >> From: Steven Ludtke <sludtke(a)bcm.edu> >> Subject: [Chimera-users] Stereo projection >> To: chimera-users(a)cgl.ucsf.edu >> Cc: mc3077(a)columbia.edu >> Message-ID: <0F963AC8-7667-4B76-B8E2-0DC9855E406C(a)bcm.edu> >> Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes >> >> Yes, we have done this before, though I don't remember with absolute >> certainty if we >> did it with chimera. You don't use 2 video cards normally, but rather >> 1 card with 2 >> outputs. Also, you will need one of the 'professional' cards where >> the >> driver supports >> stereo in a window ie- NVidia Quadro (we didn't try ATI). The NVidea >> stereo drivers then >> have a mode for this type of stereo, or at least that's my memory, it >> was a year or two >> ago when we demonstrated this. You need: >> >> 2 identical projectors which permit geometry correction so their >> corners can be exactly aligned >> 2 polarizer plates. we got (http://www.3dstereo.com/viewmaster/pj- >> pfilt-3x3.html >> ) for $20 a pair >> linear polarizing glasses (from the same site) >> Quadro or similar GFX card >> >> Note that with this type of stereo, you could technically make this >> work with any old dual >> output 3-D card, but you would have to write the visualization >> software so it would open >> 2 display windows with slightly different orientations that you could >> then manually (or >> automatically) position on each of the 2 projectors. >> >> We did this just as an initial proof of concept, and haven't managed >> to get back to it again, >> though we keep meaning to, as it is MUCH cheaper than the commercial >> solutions. Oh, one warning, >> the polarizer plates mentioned above are plastic, and the light >> energy >> put out by your typical >> projector is quite substantial. We actually melted one of the >> polarizers on our first attempt, >> despite using much of it's 3x3 inch surface area. ie - use a low >> power >> projector, or put the >> polarizer between thick glass plates or somesuch... >> >>> From: Magali Cottevieille <mc3077(a)columbia.edu> >>> Date: May 21, 2008 3:06:07 PM CDT >>> To: chimera-users(a)cgl.ucsf.edu >>> Subject: [Chimera-users] Stereo projection >>> >>> Hi, >>> >>> Is there any way, with two projectors piloted by two graphic >>> cards, to >>> send from Chimera the stereo left eye view onto one, and the stereo >>> right eye view onto the other one ? These two projectors are meant >>> to be >>> used with filters and special glasses to create a 3D view. >>> Thanks for your help! >>> >>> -- >>> Magali Cottevieille, Ph.D. >>> Howard Hughes Medical Institute, >>> Department of Biochemistry and Molecular Biophysics, >>> 630 W 168th St, P&S Black Building 2-221 >>> New York, NY 10032 >>> >>> Ph: (+1) 212-305-9521 >>> Fax: (+1) 212-305-9500 >>> >>> Email: mc3077(a)columbia.edu >>> magali.cottevieille(a)gmail.com >>> >>> _______________________________________________ >>> Chimera-users mailing list >>> Chimera-users(a)cgl.ucsf.edu >>> http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users >>> >> --------------------------------------------------------------------- >> ------- >> Steven Ludtke, PhD | Baylor College of Medicine >> sludtke(a)bcm.tmc.edu | Associate Professor & Co- >> Director >> stevel(a)alumni.caltech.edu | National Center For Macromolecular >> Imaging >> V: (713)798-9020 | Dept of Biochemistry and Mol. >> Biol. >> F: (713)798-1625 | >> | Those who Do, Are >> http://ncmi.bcm.edu/~stevel | The converse also applies >> >> >> >> >> >> ------------------------------ >> >> Message: 3 >> Date: Wed, 21 May 2008 14:25:09 -0700 (PDT) >> From: Greg Couch <gregc(a)cgl.ucsf.edu> >> Subject: Re: [Chimera-users] Stereo projection >> To: Magali Cottevieille <mc3077(a)columbia.edu> >> Cc: chimera-users(a)cgl.ucsf.edu >> Message-ID: <Pine.OSF. >> 4.63.0805211409520.1818353(a)guanine.cgl.ucsf.edu> >> Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed >> >> On Wed, 21 May 2008, Magali Cottevieille wrote: >> >>> Hi, >>> >>> Is there any way, with two projectors piloted by two graphic >>> cards, to >>> send from Chimera the stereo left eye view onto one, and the stereo >>> right eye view onto the other one ? These two projectors are >>> meant to be >>> used with filters and special glasses to create a 3D view. >>> Thanks for your help! >> >> Chimera supports stereo using workstation graphics cards, i.e., >> NVidia >> Quadro FX and ATI/AMD FireGL cards. We don't have a recent FireGL >> card >> (and driver), so the rest of this email will be about the NVidia >> Quadro FX >> cards and driver. >> >> The Quadro FX graphics cards have two video outputs and the driver >> lets >> you display the left eye on one output and the right eye on the >> other. >> The higher end cards have SLI support (on FireGL it's called >> CrossFire) >> which lets you use two matched graphics cards together. It is >> unclear to >> me how or if SLI/CrossFire and stereo work together, but I would >> expect >> them to. >> >> Greg >> >> >> ------------------------------ >> >> Message: 4 >> Date: Thu, 22 May 2008 10:51:29 -0700 >> From: Eric Pettersen <pett(a)cgl.ucsf.edu> >> Subject: Re: [Chimera-users] findhbond/5750 (FindHBond failure) >> To: Tyler Arbour <tyler.arbour(a)gmail.com> >> Cc: Chimera BB <chimera-users(a)cgl.ucsf.edu> >> Message-ID: <A4A9CFED-F357-4943-AB11-43D644C07DF0(a)cgl.ucsf.edu> >> Content-Type: text/plain; charset="us-ascii" >> >> Hi Tyler, >> The problem is that Chimera "knows" the atom types of standard >> residues and doesn't try to compute them from scratch. Therefore it >> assigns ND1 in HIS a sp2 type, which throws a monkey wrench into the >> find-hbond computation as it looks for the planarity of the ND1 and >> sees it bonded to nothing! >> For what you want to do, you want to get that atom's residue >> changed. Do this: >> >> 1) Control-double click on the ND1 atom >> 2) Chose "Modify Atom" from the resulting popup menu. >> 3) Change the "Element" to "N"; leave "Geometry" as "tetrahedral" >> 4) Change "Bonds" to 3 >> 5) In the "Residue Name" section, use the "new residue" option with >> "NH3" as the name and chain "het" >> 5) Click the Change button >> >> That should give you what you want. >> >> --Eric >> >> Eric Pettersen >> UCSF Computer Graphics Lab >> http://www.cgl.ucsf.edu >> >> On May 22, 2008, at 10:25 AM, Tyler Arbour wrote: >> >>> Wow, I didn't expect a response for this...great! I was actually >>> just trying to create a figure representing a computational model >>> that I am using. Starting from the crystal structure, I wanted to >>> actually change the histidinal N to a simple NH3 molecule. I am >>> just trying to get familiar with Chimera--I usually use DS >>> Viewerpro for modifying structures, and it doesn't seem like >>> Chimera is well-suited to this type of modification. If you have >>> any advice on an easy way to make this conversion so that I have a >>> nice figure, I'd greatly appreciate it! Thanks. >>> >>> Tyler >>> >>> On May 22, 2008, at 11:17 AM, Eric Pettersen wrote: >>> >>>> Hi Tyler, >>>> Your structure seems to have a histidine residue (residue 159 in >>>> chain A) that consists solely of an ND1 atom! How did you manage >>>> that? >>>> >>>> --Eric >>>> >>>> >>>> Eric Pettersen >>>> UCSF Computer Graphics Lab >>>> http://www.cgl.ucsf.edu >>>> >>>> >>> >> >>

1 0

Code source for Multi-touch
by Gergely Nagy 20 May '08

20 May '08

Dear Chimera Dev Team, I am referring to my email that I wrote in the beginning of April where I told about our diploma work. Just to sum it up in a couple of lines, we have a multi-touch screen (30") and we would like to make Chimera work with it. We would like to work with Chimera, because at the Center of Integrative Genomics at the University of Lausanne and at the Swiss Institute of Bioinformatics people use this software so this adaptation would even be useful besides that it would be our diploma work. So a couple of questions arise. Would it be possible to have the entire source code? We are ready to sign an NDA or any other contract if necessary. I am asking this because I am having difficulties to imagine how we could make the changes without being able to check how it works in the program. If there is a way to do it rapidly then of course we won't need the entire code source. The library that handles the touch events (touchDown, touchUpdate, touchRemove) is written in C++ (would wrap it of course). It gives basically the coordinates of the touches and updates the coordinates when dragging the finger. So we would need to replace mouse events with those and make zooming, translating and rotating functions available by using only the fingers. We would also make a new menu bar with only a couple of functions to be able to demonstrate the concept. So please tell us what would be possible and what wouldn't so we can evaluate the situation and make the right decisions immediately. Thanks a lot for taking the time to reply. Best regards, Gergely Nagy Marco Pambianchi

3 2

Error on command-line startup: OSX 10.5
by Bernard Heymann 05 May '08

05 May '08

When I start up Chimera from the command line under Leopard, I get a lot of messages such as: The process has forked and you cannot use this CoreFoundation functionality safely. You MUST exec(). Break on __THE_PROCESS_HAS_FORKED_AND_YOU_CANNOT_USE_THIS_COREFOUNDATION_FUNCTIONALITY___YOU_MUST_EXEC__ () to debug. Chimera still runs, but I thought this is probably something to be dealt with sooner than later. I did some digging and came across the following technical note from Apple: http://developer.apple.com/technotes/tn2005/tn2083.html The reference to the fork message is at the bottom of the page. I also found on another site the following advice for an X11 application: "I've updated the OSX Readme TclTk build example to include an extra configure option: --disable-corefoundation Since the example is for an X11 TclTk build, CoreFoundation isn't really necessary, so this is OK." Because Chimera uses Tk, I thought this might be something to try to avoid these error messages and possible instability down the line. Bernard Heymann Rm 1515, 50 South Dr, NIH, Bethesda, 20892-8025 Tel 301-451-8241, Fax 301-480-7629

3 2

Re: [chimera-dev] [Chimera-users] Dialog w/ sliders?
by Eric Pettersen 21 Apr '08

21 Apr '08

On Apr 21, 2008, at 12:42 PM, Randy Heiland wrote: > I was wondering if there's a dialog in the gui code that has built-in > sliders? E.g., currently our extension module subclasses > ModelessDialog; however, our input panel has so many values that it > would be convenient if it was embedded in a widget such that sliders > appeared when it was resized smaller. Hi Randy, Not directly. One possibility is to use Pmw.ScrolledFrame (http:// pmw.sourceforge.net/doc/ScrolledFrame.html) as the parent widget for the contents of your dialog. Take the frame you get as an arg to the fillInUI() method and immediately create a Pmw.ScrolledFrame with that frame as the parent. Then fill your dialog in from the frame you get from Pmw.ScrolledFrame.interior(). You can do this almost painlessly like so: def fillInUI(self, parent): sf = Pmw.ScrolledFrame(parent) sf.grid(sticky="nsew") origParent, parent = parent, sf.interior() # remainder of fillInUI() remains the same... Another possibility is to use the Feature_Buttons_Panel widget from Chimera's CGLtk.Hybrid module. This is what Volume Viewer uses to hide/show sections of its interface. You'd have to look at the Volume Viewer code to see how it's called since I haven't used it myself. --Eric Eric Pettersen UCSF Computer Graphics Lab http://www.cgl.ucsf.edu

1 0

Chimera in windows: seeing samba mounted disk
by Jean-Didier Maréchal 18 Apr '08

18 Apr '08

Dear all, I usual use chimera on my linux box, but for some practicities, I need to use it in windows. I realized though that I can not access the disk F: of my windows box which is a mounted disk from the linux one. Would it be difficult that the file browser of chimera recognize samba mounted disk? Our is ther a trick I miss? All the best, JD On Wed, 2008-04-09 at 12:00 -0700, chimera-dev-request(a)cgl.ucsf.edu wrote: > Send Chimera-dev mailing list submissions to > chimera-dev(a)cgl.ucsf.edu > > To subscribe or unsubscribe via the World Wide Web, visit > http://www.cgl.ucsf.edu/mailman/listinfo/chimera-dev > or, via email, send a message with subject or body 'help' to > chimera-dev-request(a)cgl.ucsf.edu > > You can reach the person managing the list at > chimera-dev-owner(a)cgl.ucsf.edu > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Chimera-dev digest..." > > > Today's Topics: > > 1. Re: Multi-touch and Chimera (Tom Goddard) > 2. Re: Multi-touch and Chimera (Tom Goddard) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Tue, 08 Apr 2008 12:39:58 -0700 > From: Tom Goddard <goddard(a)cgl.ucsf.edu> > Subject: Re: [chimera-dev] Multi-touch and Chimera > To: Gergely Nagy <gergelynagy76(a)gmail.com> > Cc: chimera-dev(a)cgl.ucsf.edu > Message-ID: <47FBCA0E.8090702(a)cgl.ucsf.edu> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > Hi Greg, > > Chimera is written in Python and C++. The C++ is used for OpenGL > rendering and for optimizing calculations. The event loop and user > interface is all in Python. The touch events and the actions they > trigger should definitely be in Python. It looks very reasonable to use > Boost Python to wrap the C++ touchlib library or parts of it to make it > accessible to Python. > > http://wiki.python.org/moin/boost.python/GettingStarted > > http://www.boost.org/doc/libs/1_35_0/libs/python/doc/index.html > > We use an in-house wrapping program (WrapPy). Wrapping C/C++ for Python > is often done with package called SWIG. > > We use the Tcl/Tk event loop using the Python/Tk interface module > called Tkinter. You can put custom Tk event types such as touch events > in the event queue but I haven't figured out a way to do it. The > trouble is that you typically will get an event from a callback in C/C++ > code. I recently tried passing such an event (for a Space Navigator > 6-dof input device) to a Python routine. It resulted in crashes because > Tkinter is designed so that only it can call Python code. It has some > complicated thread locking code that releases a Python lock and my event > handler called into Python without having that Python lock, tried to do > some Tk call and crashed. Probably the Python lock can be acquired in > the C/C++ callback, but then if that Python code does Tk calls to add an > event there may be additional problems with Tkinter's Tk lock which is > inaccessible. So some work is needed to figure out how to get the C/C++ > callback events into the Tk event loop. Once that is solved everything > can be done in Python with event handlers processing those Tk events to > cause any desired Chimera actions. > > Tom > > > Gergely Nagy wrote: > > Hi Tom, > > > > Thanks for your reply. > > Our multi-touch interface won't work with touch-pads, since it is > > based on another technology. > > The user will need to touch the actual object on the screen since > > there won't be any pointers. The interaction is direct between the > > user finger and the programs objects. So basically when the user > > touches a menu item for example, there is a sort of mouse clicked > > event at the given coordinates on the screen. > > Talking about events, there exists a library written in C++ called > > Touchlib (http://nuigroup.com/touchlib/) that pass those touch events > > to any program written in C++ and this is what we would like to use. > > Of course in this case we will need to wrap that code to be able to > > use it in python. I thought using Boost > > (http://www.boost.org/users/index.html) for that. > > I was asking you about the sources because at the end we should have > > something compiled and working. It doesn't matter if its not the > > latest version, what would be important is that we will be able to > > demonstrate the interface working. Would it be possible? > > > > Concerning the atom-atom distance, the user would zoom in on the part > > he is interested in, after he would choose the appropriate display > > form and then he would change from moving-zooming mode to selection > > mode and he would touch the atoms he would like to know the distance > > between. So you are right, visual feedback would be necessary, as it > > was the case with a mouse click. > > So to sum up the interaction events, there would be something like > > finger_down, finger_up, dragging. And also these events are dispatched > > by touchlib, so we would need to implement them in Chimera. > > > > What do you think about it? > > > > Thanks a lot for your help! > > > > Regards, > > Greg > > > > > > > > 2008/4/8, Tom Goddard <goddard(a)cgl.ucsf.edu > > <mailto:goddard@cgl.ucsf.edu>>: > > > > Hi Greg, > > > > Sounds neat. Will a multi-touch interface work with a standard > > laptop touch-pad? > > > > The source code for Chimera is on the web though it isn't up to date. > > > > http://www.cgl.ucsf.edu/chimera/sourcecode.html > > > > All of the functions you are interested in are implemented in > > Python code which is included in the distributed Chimera. The > > Python code is in files (with ".py" suffix) in directories under > > > > chimera/share > > > > or on Mac under directory Chimera.app/Contents/Resources/share. > > But it is quite difficult to find the Chimera Python routines you > > need. Our programming reference guide (automatically generated > > from code comments) is not too useful but we provide programming > > examples that might help. > > > > > > http://www.cgl.ucsf.edu/chimera/docs/ProgrammersGuide/Examples/index.html > > > > The easiest approach is to ask us by email for the routines you > > need, e.g. list all chain identifiers, select chain given id, > > color equivalent to actions menu, move models, rotate models, show > > atom-atom distance, find atom under mouse, draw label.... Eric > > Pettersen is the most familiar with that code and what simple > > routines already exist to do those things. He is on vacation now > > but expected back around Thursday. > > > > You are interested in atom-atom distances. Will the user choose > > an atom through the touch interface? It seems that would require > > visual feedback in the Chimera window -- equivalent of the mouse > > pointer. I suppose you could take over the mouse pointer, or you > > could create a new "cursor" implemented perhaps as a Chimera 2d label. > > > > Tom > > > > > > > > > > > > > ------------------------------ > > Message: 2 > Date: Tue, 08 Apr 2008 12:49:01 -0700 > From: Tom Goddard <goddard(a)cgl.ucsf.edu> > Subject: Re: [chimera-dev] Multi-touch and Chimera > To: Gergely Nagy <gergelynagy76(a)gmail.com> > Cc: chimera-dev(a)cgl.ucsf.edu > Message-ID: <47FBCC2D.2010402(a)cgl.ucsf.edu> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > Greg suggest a direct call from C/C++ to queue an event. > > Tom > > Greg Couch wrote: > > Tom, > > > > You need to link with tcl (actually tcl stubs) and use the > > Tcl_ThreadQueueEvent function. > > > > - Greg > > > > On Tue, 8 Apr 2008, Tom Goddard wrote: > > > >> From: Tom Goddard <goddard(a)cgl.ucsf.edu> > >> Sender: chimera-dev-bounces(a)cgl.ucsf.edu > >> To: Gergely Nagy <gergelynagy76(a)gmail.com> > >> Cc: chimera-dev(a)cgl.ucsf.edu > >> Date: Tue, 08 Apr 2008 12:39:58 -0700 > >> Subject: Re: [chimera-dev] Multi-touch and Chimera > >> Received-SPF: pass (cgl.ucsf.edu: 169.230.27.3 is authenticated by a > >> trusted > >> mechanism) > >> Received-SPF: pass (cgl.ucsf.edu: 169.230.27.3 is authenticated by a > >> trusted > >> mechanism) > >> > >> Hi Greg, > >> > >> Chimera is written in Python and C++. The C++ is used for OpenGL > >> rendering and for optimizing calculations. The event loop and user > >> interface is all in Python. The touch events and the actions they > >> trigger should definitely be in Python. It looks very reasonable to use > >> Boost Python to wrap the C++ touchlib library or parts of it to make it > >> accessible to Python. > >> > >> http://wiki.python.org/moin/boost.python/GettingStarted > >> > >> http://www.boost.org/doc/libs/1_35_0/libs/python/doc/index.html > >> > >> We use an in-house wrapping program (WrapPy). Wrapping C/C++ for Python > >> is often done with package called SWIG. > >> > >> We use the Tcl/Tk event loop using the Python/Tk interface module > >> called Tkinter. You can put custom Tk event types such as touch events > >> in the event queue but I haven't figured out a way to do it. The > >> trouble is that you typically will get an event from a callback in C/C++ > >> code. I recently tried passing such an event (for a Space Navigator > >> 6-dof input device) to a Python routine. It resulted in crashes because > >> Tkinter is designed so that only it can call Python code. It has some > >> complicated thread locking code that releases a Python lock and my event > >> handler called into Python without having that Python lock, tried to do > >> some Tk call and crashed. Probably the Python lock can be acquired in > >> the C/C++ callback, but then if that Python code does Tk calls to add an > >> event there may be additional problems with Tkinter's Tk lock which is > >> inaccessible. So some work is needed to figure out how to get the C/C++ > >> callback events into the Tk event loop. Once that is solved everything > >> can be done in Python with event handlers processing those Tk events to > >> cause any desired Chimera actions. > >> > >> Tom > > > > ------------------------------ > > _______________________________________________ > Chimera-dev mailing list > Chimera-dev(a)cgl.ucsf.edu > http://www.cgl.ucsf.edu/mailman/listinfo/chimera-dev > > > End of Chimera-dev Digest, Vol 51, Issue 3 > ******************************************

2 1

Re: [chimera-dev] Multi-touch and Chimera
by Tom Goddard 08 Apr '08

08 Apr '08

Hi Greg, Chimera is written in Python and C++. The C++ is used for OpenGL rendering and for optimizing calculations. The event loop and user interface is all in Python. The touch events and the actions they trigger should definitely be in Python. It looks very reasonable to use Boost Python to wrap the C++ touchlib library or parts of it to make it accessible to Python. http://wiki.python.org/moin/boost.python/GettingStarted http://www.boost.org/doc/libs/1_35_0/libs/python/doc/index.html We use an in-house wrapping program (WrapPy). Wrapping C/C++ for Python is often done with package called SWIG. We use the Tcl/Tk event loop using the Python/Tk interface module called Tkinter. You can put custom Tk event types such as touch events in the event queue but I haven't figured out a way to do it. The trouble is that you typically will get an event from a callback in C/C++ code. I recently tried passing such an event (for a Space Navigator 6-dof input device) to a Python routine. It resulted in crashes because Tkinter is designed so that only it can call Python code. It has some complicated thread locking code that releases a Python lock and my event handler called into Python without having that Python lock, tried to do some Tk call and crashed. Probably the Python lock can be acquired in the C/C++ callback, but then if that Python code does Tk calls to add an event there may be additional problems with Tkinter's Tk lock which is inaccessible. So some work is needed to figure out how to get the C/C++ callback events into the Tk event loop. Once that is solved everything can be done in Python with event handlers processing those Tk events to cause any desired Chimera actions. Tom Gergely Nagy wrote: > Hi Tom, > > Thanks for your reply. > Our multi-touch interface won't work with touch-pads, since it is > based on another technology. > The user will need to touch the actual object on the screen since > there won't be any pointers. The interaction is direct between the > user finger and the programs objects. So basically when the user > touches a menu item for example, there is a sort of mouse clicked > event at the given coordinates on the screen. > Talking about events, there exists a library written in C++ called > Touchlib (http://nuigroup.com/touchlib/) that pass those touch events > to any program written in C++ and this is what we would like to use. > Of course in this case we will need to wrap that code to be able to > use it in python. I thought using Boost > (http://www.boost.org/users/index.html) for that. > I was asking you about the sources because at the end we should have > something compiled and working. It doesn't matter if its not the > latest version, what would be important is that we will be able to > demonstrate the interface working. Would it be possible? > > Concerning the atom-atom distance, the user would zoom in on the part > he is interested in, after he would choose the appropriate display > form and then he would change from moving-zooming mode to selection > mode and he would touch the atoms he would like to know the distance > between. So you are right, visual feedback would be necessary, as it > was the case with a mouse click. > So to sum up the interaction events, there would be something like > finger_down, finger_up, dragging. And also these events are dispatched > by touchlib, so we would need to implement them in Chimera. > > What do you think about it? > > Thanks a lot for your help! > > Regards, > Greg > > > > 2008/4/8, Tom Goddard <goddard(a)cgl.ucsf.edu > <mailto:goddard@cgl.ucsf.edu>>: > > Hi Greg, > > Sounds neat. Will a multi-touch interface work with a standard > laptop touch-pad? > > The source code for Chimera is on the web though it isn't up to date. > > http://www.cgl.ucsf.edu/chimera/sourcecode.html > > All of the functions you are interested in are implemented in > Python code which is included in the distributed Chimera. The > Python code is in files (with ".py" suffix) in directories under > > chimera/share > > or on Mac under directory Chimera.app/Contents/Resources/share. > But it is quite difficult to find the Chimera Python routines you > need. Our programming reference guide (automatically generated > from code comments) is not too useful but we provide programming > examples that might help. > > > http://www.cgl.ucsf.edu/chimera/docs/ProgrammersGuide/Examples/index.html > > The easiest approach is to ask us by email for the routines you > need, e.g. list all chain identifiers, select chain given id, > color equivalent to actions menu, move models, rotate models, show > atom-atom distance, find atom under mouse, draw label.... Eric > Pettersen is the most familiar with that code and what simple > routines already exist to do those things. He is on vacation now > but expected back around Thursday. > > You are interested in atom-atom distances. Will the user choose > an atom through the touch interface? It seems that would require > visual feedback in the Chimera window -- equivalent of the mouse > pointer. I suppose you could take over the mouse pointer, or you > could create a new "cursor" implemented perhaps as a Chimera 2d label. > > Tom > > > >

1 1

Multi-touch and Chimera
by Gergely Nagy 08 Apr '08

08 Apr '08

Hello all, I am a graduate student at the University of Applied Sciences in Switzerland and in my diploma work, I would like to develop a multi-touch user interface for Chimera. By multi-touch I mean the capability of interpreting simultaneous user interactions. You can find more info on the technology here: http://cs.nyu.edu/~jhan/ftirtouch/ <http://cs.nyu.edu/%7Ejhan/ftirtouch/> I already have a working hardware, now I have to find the software to work on. Of course the time and resource limitations of this diploma work wouldn't allow to interface all the functions of Chimera for multi-touch, so I'd be aiming to choose a couple of them, those that are used mainly by the users of Chimera that I know. So basically my aim would be to take for example the Select-chain menu, the Actions-color menu and the Tools-Structure Analysis-Distances tool and make it available to touch interaction. And of course the user would be able to move around the molecule with one finger along the X and Y axis, he would be able to turn it around using 3 fingers simultaneously (2 defining the axis and the third making the rotation) and zoom in and out using 2 fingers simultaneously. So my question is, would it be possible to have the source code that would let me do this development? Thanks a lot for taking the time to read my inquiry. Regards, Greg (Gergely Nagy)

2 1

Re: [chimera-dev] [Chimera-users] list of atoms attributes
by Eric Pettersen 01 Apr '08

01 Apr '08

On Feb 22, 2008, at 7:28 AM, Jean-Didier Maréchal wrote: > Hi again, > > When using the python shell, I am often struggling to find the > attribute for the atoms I work on. > for at my list with the selected atoms, I caught that at[i].name gives > the atom name of the entry at[i], at[i].residue provides with the name > residue.chain, etc. > > Where can I find in the programmer's guide the full list of available > attributes? What would be the way to ask for the serialNumber from the > IDLE? > > Thanks lot > JD Hi JD, dir(a) will list the attributes of 'a'. help(a) will produce documentation for the attributes of 'a' that come from the C++ layer (but not ones that were added in the Python layer). 'a' could be a class rather than an instance. For example, help(chimera.Atom) lists the methods/attributes of Atom and provides some info about them. We do what we can to keep the programmer's guide up to date, but for attributes (since they get added/changed pretty frequently) it is better to query with help()/dir() than to fall back on the relatively static programmer's guide documentation. --Eric Eric Pettersen UCSF Computer Graphics Lab http://www.cgl.ucsf.edu

3 2

Wrapping Chimera in a Java program
by Fabrice Jossinet 31 Mar '08

31 Mar '08

Dear developers, I would like to launch Chimera from my Java program. I would like to send messages to Chimera to highlight in its own display selections I do in my Java program's display. I already did that with PyMOL using a pipe communication. So, several questions: - is it possible to start Chimera from a classical python distribution (not the embedded python of Chimera)? One of my ideas is to start Chimera from Jython, a Java implementation of Python - is it possible to communicate with Chimera using a pipe ? - and the most general question: is it possible to communicate with Chimera from the outside? Thank you for your answers Fabrice -- Dr. Fabrice Jossinet Laboratoire de Bioinformatique, modelisation et simulation des acides nucleiques Universite Louis Pasteur Institut de biologie moleculaire et cellulaire du CNRS UPR9002, Architecture et Reactivite de l'ARN 15 rue Rene Descartes F-67084 Strasbourg Cedex France Tel + 33 (0) 3 88 417053 FAX + 33 (0) 3 88 60 22 18 f.jossinet(a)ibmc.u-strasbg.fr fjossinet(a)gmail.com http://www-ibmc.u-strasbg.fr/arn/Westhof/index.html http://fjossinet.u-strasbg.fr/

2 1

new keyword arg for added commands
by Eric Pettersen 26 Mar '08

26 Mar '08

If you have code that adds commands to Chimera using the addCommand () function, there is a new keyword to that function so that animation scripts can work better. The keyword is 'changesDisplay' and is a boolean that defaults to True. If your command typically changes the display in the main graphics window somehow the keyword should be True, otherwise False. It's needed so that as a script is executed, Chimera knows after each line whether a new frame is necessary or not. For instance, for the 'alias' command the value is False since after executing that command in a script you would not want to record a new frame in an animation. --Eric Eric Pettersen UCSF Computer Graphics Lab http://www.cgl.ucsf.edu

1 0

fit map in map -- also fit a/pix at the same time ?
by stefan.backstrom＠medchem.umu.se 26 Mar '08

26 Mar '08

Hi. Would it be possible to optimize the a/pix, or scale value when one is fitting either a map in map, or model in map ? Often it seems as if the scale is not 100 % correct between a crystal structure and a cryo-EM reconstructinon, and it would be nice to be able to choose to fit the scale of the EM-map at the same time as the fit rotation and translation. Best wishes, Stefan Bäckström Medical Biochemistry and Biophysics Umeå University Sweden

2 1

iPhone SDK
by Tom Ferrin 11 Mar '08

11 Mar '08

The iPhone SDK was announced last week. Turns out the iPhone includes a lot of the same OS layers as does the Mac. I didn't see anything about Python, but there is a hardware accelerated OpenGL ES in there. Also the standard set of development tools (interface builder, xcode, etc) and a package called "Instruments" that let's you capture on a Mac lots of execution trace information on the iPhone via an interface cable. One novel user interface component: the iPhone includes a 3D accelerometer with software interface. The video I watched showed a shoot-em-up space wars game where you steered by moving the phone around. Someone is bound to come up with an interactive molecular viewer; I suspect Warren DeLano is working on Pymol right now. --tom

2 1

Re: [chimera-dev] [Chimera-users] Driving matchAtoms from the IDLE
by Eric Pettersen 22 Feb '08

22 Feb '08

On Feb 22, 2008, at 3:45 AM, Jean-Didier Maréchal wrote: > Hi Guys, > > I'd like to superpose two inorganic complexes using two differently > "ordered" atom sets. My compound has quite similar substructures but I > am sure at all what is the best "ordered" selection of atom pairs to > align. Therefore, I'd like to make a "extensive" search (with an > heuristic approach to define...) passing by the idle. I am > unfortunately > missing something before getting to this point! > > Lets say that I just want to align the nitrogen atoms of the my two > structures. I have: > > import chimera > import numpy > from chimera import match > from chimera import runCommand > runCommand("select #1@N=") > an1=chimera.selection.currentAtoms() > runCommand("select #0@N=) > an2=chimera.selection.currentAtoms() > chimera.match.matchAtoms(an1,an2) > > And then it's when I am already stuck! > > the resulting prompt does not provide mistake an gives: > chimera.match.matchAtoms(an1,an2) > (<_chimera.Xform object at 0x9463ef8>, 0.034256915079603358) > > But how this results can be send to the graphical window? I guess I > have > to use the Xform object, but I can not clearly understand how to do > this. > > Thanks for any help! > > JD Hi JD, I'm not 100% certain what your goal is, but you could just run the match command directly: from chimera import runCommand runCommand("match #1@N= #0@N=") If you need access to the RMSD of the match, you could call the underlying match() function: from Midas import match mobileAtoms, refAtoms, rmsd = match("#0@N=", "#1@N=") which will also superimpose the structures. --Eric Eric Pettersen UCSF Computer Graphics Lab http://www.cgl.ucsf.edu

1 0

change hand ?
by stefan.backstrom＠medchem.umu.se 15 Feb '08

15 Feb '08

Hi. Is it possible to flip the hand of a cryo-EM map in chimera ? Best wishes, Stefan Bäckström Medical Biochemistry and Biophysics Umeå University Sweden

1 0

Chimera batch surface exporter
by Trent Grover 23 Jan '08

23 Jan '08

Hi Everyone, I'm pretty new to Chimera, but I'm working on an animation project that requires quite a few proteins and molecules. To speed up the process, I'd like to create a batch converter/exporter script to do the following: Open a PDB file Generate a multiscale model at a specified resolution Export the model out as a VRML or OBJ file I think I can do this using a Chimera command file, but I was hoping someone could point me in the right direction for some of the relevant command line syntax. Is it possible to control the multiscale model tool via command line (I know I can launch the dialog, but can I change the parameters and execute)? Is there a reference somewhere for the command line parameters for the multiscale models tool? What is the command line export command? Where can I find an OBJ exporter (there are references to it in the mailing list archive)? Any help would be greatly appreciated. Thanks, Trent BTW - I have a suggestion for improvement of Chimera. Add a command history window (or modify the existing one) that echoes out the command line text that is equivalent to any command executed via the Chimera user interface. Maya has this functionality and it is wonderful for anyone who is trying to figure out how to script a process that they know how to execute via the various menus.

1 0

saving custom molecular attributes
by Eric Pettersen 14 Jan '08

14 Jan '08

If you are writing any code that adds non-standard attributes to Atoms/Molecules/Residues/Bonds and you want those attributes preserved in sessions, you can now use the registerAttribute() function to do this. I've appended more info below. --Eric Eric Pettersen UCSF Computer Graphics Lab http://www.cgl.ucsf.edu Custom Molecular Attributes If you add non-standard attributes to Molecules/Residues/Bonds/Atoms that you want preserved in restored sessions, use SimpleSession's restoreAttribute() function, e.g,: import chimera from SimpleSession import registerAttribute registerAttribute(chimera.Molecule, "qsarVal") Note that the Define Attribute tool automatically does this registration, so it's only attributes that you add directly from your own Python code that need to be registered as above. Also, only attributes whose values are recoverable from their repr() can be saved by this mechanism, so values that are C++ types (Atoms, MaterialColors, etc.) could not be preserved this way.

1 0

restoring hidden models
by Eric Pettersen 11 Jan '08

11 Jan '08

I just changed the session save/restore code to process hidden Molecular models also (need it for Rotamers session support). You can still prevent such models of your own from being processed with the SimpleSession.noAutoRestore(model) function. --Eric Eric Pettersen UCSF Computer Graphics Lab http://www.cgl.ucsf.edu

1 0

SimpleSession.sesRepr()
by Eric Pettersen 20 Dec '07

20 Dec '07

Hi all, For anyone saving data into session files, I've implemented a sesRepr () function for use in place of repr() for large data objects (ones whose repr() would be hundreds or thousands of characters). sesRepr () inserts newlines periodically [where possible] so that the resulting session file has a better chance of being viewable in an editor or sent through various mailers without corruption. sesRepr() will also frequently make the representation smaller despite the additional newlines since it passes float values through the "%g" format which will often trim their representation by half a dozen characters or so. I thought pprint.pformat() would do the job but it adds a _lot_ of whitespace to the repr if there are nested compound objects (e.g. a dictionary whose values are lists). You access sesRepr() (as of tomorrow's daily build) with: from SimpleSession import sesRepr --Eric Eric Pettersen UCSF Computer Graphics Lab http://www.cgl.ucsf.edu

1 0

modified chimera to use an unmodified Python
by Greg Couch 19 Dec '07

19 Dec '07

It will be available in tomorrow's daily builds. So in principle, it will be possible to run chimera inside another Python program by doing "import chimeraInit; chimeraInit.init(eventloop=False)". You should also incorporate into the other program's Python library the bug fixes chimera has in its Python library. This is a development version of chimera and might have major bugs in it. If you get any tracebacks that have the word "metatype" in them, please send them to me ASAP. - Greg

1 0

Re: [chimera-dev] printing "zone" search
by Eric Pettersen 14 Dec '07

14 Dec '07

Hi Ryan, The difference is that when you print a list, the residues in the list are shown with repr(). When you print a residue directly, it is shown with str(). Quoting the Python Tutorial, the difference between str() and repr() is: "The str() function is meant to return representations of values which are fairly human-readable, while repr() is meant to generate representations which can be read by the interpreter..." Since there is no representation of a Residue object that can be directly read by the interpreter, you get the output you saw when printing the list. The result of str() when you directly print the residue is controlled by the "Atomspec display style" preference in the General preferences category (Favorites->Preferences). It seems you have that preference set to "command-line specifier". If you change it to "simple" instead you will get output like "LYS 12.A" etc. Alternatively you could use the chimeraLabel function, e.g.: print chimera.misc.chimeraLabel(r, style="simple") Programmer documentation for Chimera is considerably less complete than user documentation. My best advice is to read through some of the Programmer's Examples (http://www.cgl.ucsf.edu/chimera/docs/ProgrammersGuide/Examples/ index.html) and look through the Programmer's FAQ (http://www.cgl.ucsf.edu/chimera/docs/ProgrammersGuide/faq.html) If you know that some Chimera tool does something similar to what you want then browse through its source code. If you don't know Python that well you might want to work through the Python Tutorial if you have time (http://www.python.org/doc/current/tut/tut.html) And feel free to ask questions; we're happy to help. --Eric On Dec 14, 2007, at 4:55 PM, Ryan wrote: > Hi Eric, thanks for your help. I'm getting closer to achieving the > results i'd like. > > So i've been able to properly select the residues within a specified > radius of my ligand, but am a little confused as to how Chimera > handles the selection list. If i try: > > print chimera.selection.currentResidues() > > i get a list like this: > >> [<_chimera.Residue object at 0x578d2020>, <_chimera.Residue object at >> 0x578cbfb0>, ... ] > > but if i try: > > for x in chimera.selection.currentResidues(): > print x > > i get output that is much closer to what i'm looking for: > >> #0:499 >> #0:496 >> ... > > What is the reason for the difference between these two forms of > output that seem to be accessing the same list? Also, is there a way > i can get more detailed output with residue names and numbers, such as > ['#0:TYR499','#0:TRP496',...]? > > Sorry if these questions are answered elsewhere, but i could not find > the answers online. Is there a document out there that can teach me > how to better master the Chimera Python modules for scripting? I'm no > Python expert -- i've so far only been learning things as i need them. > Thanks again, > > ryan >

1 0

Re: [chimera-dev] [Chimera-users] printing "zone" search
by Eric Pettersen 14 Dec '07

14 Dec '07

On Dec 14, 2007, at 8:23 AM, Ryan wrote: > I am actually currently interested in doing something similar to > Francesco's previous question about writing out a list of residue > names within a certain zone of a ligand -- however i would like to > do this from a Python script. > > I tried to find the zone selection command to use in a script, but > could not. Would it be easiest to just use the runCommand('z<') > function to select a zone? Using runCommand() would be by far the easiest and is what I would recommend. Nonetheless, there are alternatives. I've appended a script that finds all NZ atoms of lysines that are within 8 angstroms of an NZ of a different lysine. It calls into the underlying chimera.specifier.zone() function since it needs to do additional processing to exclude each lysine's own NZ atom when examining the zones for NZs within 8 angstroms. If efficiency is important (and it doesn't sound like it really is from your description) you would want to consider using the _closepoints module to quickly eliminate most atoms that aren't in the zone, but then you have to process the remainder to ensure that they meet the zone criteria (some remaining ones won't). Look at the start of the detectClash() function in DetectClash/__init__.py for an example. > Also, when the zone is selected, are the selected residue names > stored in a list? If so, how can i access this list to print out > the residue names into an output file of my own? chimera.selection.currentResidues() returns such a list. --Eric

1 1