Chart Homepage

Chart-0.6.0 released 26th November 2001

• chart-0.6.0.tar.gz Latest Chart release.
• mapview-0.4.3.tar.gz Latest Mapview release
• MapView has been renamed Coot and development has move to Coot's home in York.

MapView: a Simple Map and Molecule Viewer

A quick and easy way to see the results of your refinement:

MapView reads (currently) CCP4 Maps and Datasets (mtz files) and pdb and cif files and displays them in the convential chickenwire and stick representation.

There are also functions to manipulate (fit) the coordinates.

Chart: a Structure Solution Wrapper for SHELXD, SHELXS and The CCP4 Program Suite

Chart is a program that uses SHELXS, SHELXD (if you have them) and the CCP4 Program Suite to solve structure with minimum effort (on the crystallographers part).

The idea is to take the MTZ output of TRUNCATE (or the SCALEPACK .sca files) for the native and possible derivatives and run through the appropriate steps, the final one being DM (or ARP/wARP).

This means that you can routinely convert you scalepack files to coordinates in about an a few hours (with a fast computer). Various post-DM maps are created which can be examined with a crystallographic model building program (but that's another story).

• Chart is still being developed.
• I don't guarantee that Chart will solve your structure (even if it's solvable) - it should do interesting things however.

Which you may or may not already have:

• STk:

  • STk RedHat rpms (version 4.0.1)

  • STK SGI IRIX 6.5 Binary tar

  • STk-4.0.1 source code

• scsh:

  • scsh RedHat rpm

  • scsh SGI IRIX 6.5 Binary tar

  • scsh-0.5.2 source code

Note that the SGI tar files will extract into /usr/local, you may need to be root to do this.

My notes on compiling scsh and STk.

• You can see a CCP4 newsletter article here. Things have moved on a bit since that was written, obviously.

• You can read a seminar handout from the EMBO Grenoble Automation course (March 00) here.

• Chart-0.4.3 User Guide (gziped PostScript (which your browser may unzip on the fly)).

• Released 16 May 2000.
• I had forgotten to include pre-mad.scsh (I only found out yesterday when I tried to use the distributed version myself). It is included now.
• Added (more) support for ccp4i.
• DM solvent optimization typo fixed
• Added special position filtering of Se atoms in MIR experiment (hehe).
• (And also a MAD experiment (less bizarre).)
• Changed Rantan weights.

• Released 24 April 2000.
• This is really 0.5.0-pre-1 i.e 0.5.0 is coming in a few days. This is here to test for bugs.
• Many bug fixes and improvements.
• Added SAD and MAD+Native. Hooray! (This is what took the time, actually).

  Although SAD has been enabled, I have not been been able to solve a structure using it. Do you have SAD data that you know provides a nice map?

  Similarly, MAD+Native has been implemented, but I don't have such a data set on which to test it ;-).

• Fixed bug in mlphare parameter refinement placement.
• Don't allow SHELX to search for Selinums on special positions.
• Add non-crystallographic-guess-at-foldedness to SHELX input.
• Various updates to scsh-0.5.2.
• Reduced MAD parameter refinement - speeding things up a bit.
• Added Rantan run using modified Fa's from remote.
• Changed freerflag input to generate better number of freeR reflections.
• Much ARP/wARP testing. I recommend always use WARP PHASED (that's "Warp Control-> DM phased" in the GUI).
• WARNING: [Added 25 April 2000] Well, that didn't take long. There is indeed an error in 0.4.3, specifically, The first MAD shelx is broken.

• Large step along the way to 0.5.0. Some warp related typos fixed.
• The CD-ROM on my computer is broken. This has speeded Chart deveopment considerably.
• Scalepack conversion improved.

• Fixed a couple of typos in the wARP parameter file.

• Released 4 Feb 2000.
• wARP code cleaned up - but not fully tested.
• MAD atom parameter fixing for strange cases.
• Updated documentation.

• Released chart-0.4.0-pre2. wARP is still not working - but some other things are fixed.

• Squished the pop-up bug.
• DM and resolution optimization work nicely now from the GUI.
• Temperature factors are recovered.
• Been doing Christmas & New Years & CCP4 Study Weekend stuff, so Chart development stalled for several weeks.
• Now that Victor and Tassos have released their GNU/Linux binaries, I have been adding in wARP support.

• Adding in optimization control to the GUI and connecting it up. There is a bug in the pop-up dismissal that I haven't managed to squish. Will wrestle with it later.
• Tidying up DM solvent optimization.
• Improving MAD mlphare initial values. Why are the temperature factors getting lost?
• Adding in ARP/wARP control to GUI.

• Released 2 Dec 1999.
• Fixes warp-related typos

• Released 2 Dec 1999.
• Fixed shelx atom type bug.
• Started adding wARP support.

• Released 30 Nov 1999.
• Added parameter manipulation to the GUI. A bit crude at the moment.
• chart.scm now includes all the user changable parameters. These should all be gui changable too in future.
• Added work-around for mtz2various bug.
• Auto-calculation of "awk" program, depending on OS.
• Fixed bug in (acceptable-number-of-centrics).
• Shuffled around startup parameterization.
• Changed to "Figure of Merit" in DM analysis.
• Added Special Position code.
• Rantan now output top 5 phase sets.
• Added "kill" button.

• Adding support for scalepack ".sca" files.
• Adding in GUI parameters.

• Column labelling fixed again.

• chart-0.3.0 released 26th Oct 1999.
• chart-0.3.0 now requires scsh-0.5.2 and (may require) STk-4.0.1 (i.e. earlier versions of STk may work).
• Atom Labels bug fixed.
• Occupancy refinement "improved".
• End tag: MAD/MIR reading fixed.

• chart-0.2.9 is a bug fix version.
• Showstopping MIR(AS) labelling bug fixed.
• The same "Atom type" is now allowed for different derivative datasets.
• Many mlphare HKLOUT files are now deleted and the best of the initial atom sets is now re-run. Much disk space saved.

• Bug fix version: There was a horrible bug in 0.2.7 that I forgot to remove in the job submission part of the GUI. Easily fixed.
• New (improved?) File box.
• final-remarks no longer thinks all MIR structure solutions fail.
• BUG: The filebox still is not quite right.
• BUG: A showstopping MIR(AS) MTZ-label bug! Fixed in 3.0 (out soon).

• Updated Rantan input (now with END cards)
• Can now run user code in ~/.chartrc
• Better contour levels
• Fixed aesoteric setup file parsing bug

• As of CCP4 version 3.5, mtz2various produces F-squareds. Change shelx input to reflect that.
• Fixes re-run-first-best "positional refinement of fixed 0.0 occupancy atom" problem.

• Released 13th Septh 1999.
• Having chart used in anger was a sobering (but useful) experience.
• Many MIR-related bug fixes.
• Still a shelx-related peculiarity under investigation.
• MAD additional atoms improved.

• There is now a preliminary User Guide.
• Better parameter refinement as solution progresses.
• Shelx solutions are no longer ignored (oops :).
• GUI "Load file" "label already exists" problem fixed. However, you still have to press "Return" on the mtz file entry widgets for the column labels to get re-read (this is on the bug list).
• Add heavy sites to final map (hooray, at last).
• Known Bug: additional-atoms map+peaksearching is still not working properly.

• To be released 7th Sept 99.
• Aesthetic bugs fixed (as far as I can tell).
• Progress Bar (really) enabled.
• Atoms plotted on map now.

• Chart-0.2.3 released 31st August 1999.
• OK, enough delay - out with it. This is not quite all I intended - particularly there are aesthetic bugs in the GUI (which are now top of the list to be fixed in 0.2.4). However, there are enough new features and bug fixes to be a considerable improvement to 0.2.2.
• Known Bug: the GUI says "Job terminated" very shortly after it has started. However, the job has not terminated (as you can see by manually tailing the chart.log file). This will be fixed in 0.2.4.
• A final remarks routine tells you if Chart thinks you have solved the structure ;-).
• Fixed B-factor reading on some refined atoms.
• More robust trial atom combination.
• Symmetry checking bug fixed (again!). Hopefully this bug is completely dead and buried.
• Set the inflection point data set occupancies to 0.0 and don't refine them.
• Delete interfering old scaleit log files before running new ones.
• Reject atoms that have a refined temperature factor greater than 200.0 (this is a user changable parameter - set in chart.scsh).
• Removed pointless and time wasting MAD single additional atom refinement (this may need to be done for MIR too).
• Sign of density consistancy of anomalous MAD sites. If they are consistent, Chart tells you it's a good sign.
• Code added to move towared log, debug and resume files (basically different levels of debugging).
• Normal probability summary added. Find it in the revise directory.

• Not released yet (provisional release date: 1st Sept)
• MAD atom rejection improved (reject only from remote and apply those rejections to other datastes).
• Fixed the symmetry rejection bug (at long last!)
• More Bug-fixes
• Progress Bar should be enabled
• Normal analysis summary

• Released: 2nd August 1999.
• Bug: You may have to mkdir $CCP4_SCR/chart/mad-mlphare/additional-atoms
• Discoved a bug in reading saved chart parameter files. Press "MIRAS" then "pure MAD" button to make things look better.
• A contour of the post-RANTAN map is produced as a popup.
• SCALEIT is wrongly executed before using MAD shelx. It produces the message "scaleit failed. That's bad!". It's not bad at all - this scaleit is irrelavent.
• If you are coming to the IUCR, do come and say hello.

• I have been incorporating the minor site addition for mad structures and had found a error in my thinking with regard to derivative numbering and labelling. I have been fixing this recently. This and other IUCR business has delayed the next version of Chart. Sigh.
• Many Bug fixes
• Sometimes Rantan cannot find the sites. Anomalous Shelx has to be used instead. This is being fixed and tested.
• Improved hand checking.
• Quick MAD hand checking (from anomalous).
• Progress Bar.

• chart-0.1.3 released on 4th June 1999.
• 2nd June 1999: Big Day... chart-0.1.3 solves (i.e. provides an interpretable map) of the chaperonin apical domain (1srv) (whatever that is) MAD structure! Woop! (thank you Martin Walsh). Takes 15 minutes.
• Hand checking and report enabled.

• chart-0.1.2 was release on June 1st 1999.

• chart-0.0.9 was released on May 25th 1999.
• Version 0.1.2 with improved MAD support will be available around the beginning of June.

• Hand checking added now. May need to be further tested.
• MAD support added. Is not the final algorithm, but it does interesting things.

• Hopefully the disk space requirement has been reduced
• The hand checking is wrong

• Make sure you have lots of disk space in the scratch directory. (Many mlphare mtz files are created (due to concurrency issues), many of which are not necessary to keep. They should be deleted after they are written. This will be fixed in 0.0.8.
• DM keywords are not optimal for the current (2.0.3) version of DM.

• Fixes the SHELX_EXE typo

• MAD analysis is not included. It's on it's way.

Behind the scences Chart has pretty much become guile-powered - largely due to the work of Gary Houston. The means that Chart for alphas may be soon available.

• Make sure that you have collected the inflection point. It's easy to miss it. Rescanning the fluorescence and recollecting if necessary is a good idea. Make sure that the inflection point data set is as complete as possible.
• If collect 3 wavelength MAD (inflection point, peak and remote), make sure that that remote is a low wavelength (higher energy) so that you get some anomalous differences from it. Collecting a high wavelength (low energy) remote is probably not worth the effort. In my opinion, only collect high wavelength (low energy) remote after you have colleted the other 3.
• Probably the best advice is go to a good beamline in the first place.

So the lesson of EMACS was, pick a good programming language to write your editor in, and your extensibility language really ought to be designed to be a good programming language. You shouldn't throw away what you know about programming language design. Well, obviously the best programming [language] is LISP. -- Richard Stallman