--- The Detailed Node Listing ---
Introduction
Mousing and Keyboarding
General Features
Scripting
Python
Backups and Undo
Coordinate-Related Features
Modelling and Building
Regularization and Real Space Refinement
Map-Related Features
Validation
Hints and Usage Tips
Other Programs
This document is the Coot User Manual, giving an overview of the interactive features. Other documentation includes the Coot Reference Manual and the Coot Tutorial. These documents should be distributed with the source code.
If have found this software to be useful, you are requested (if appropriate) to cite:
"Coot: model-building tools for molecular graphics" Emsley P, Cowtan K Acta Crystallographica Section D-Biological Crystallography 60: 2126-2132 Part 12 Sp. Iss. 1 DEC 2004
The reference for the REFMAC5 Dictionary is:
REFMAC5 dictionary: "Organization of Prior Chemical Knowledge and Guidelines for its Use" Vagin AA, Steiner RA, Lebedev AA, Potterton L, McNicholas S Long F, Murshudov GN Acta Crystallographica Section D-Biological Crystallography 60: 2184-2195 Part 12 Sp. Iss. 1 DEC 2004"
If using "SSM Superposition", please cite:
"Secondary-structure matching (SSM), a new tool for fast protein structure alignment in three dimensions" Krissinel E, Henrick K Acta Crystallographica Section D-Biological Crystallography 60: 2256-2268 Part 12 Sp. Iss. 1 DEC 2004
The reference for the the Electron Density Server is:
GJ Kleywegt, MR Harris, JY Zou, TC Taylor, A Wählby, TA Jones (2004), "The Uppsala Electron-Density Server", Acta Crystallographica Section D-Biological Crystallography 60, 2240-2249.Please also cite the primary literature for the received structures.
Coot is a molecular graphics application. Its primary focus is crystallographic macromolecular model-building and manipulation rather than representation i.e. more like Frodo than Rasmol. Having said that, Coot can work with small molecule (SHELXL) and electron microscopy data, be used for homology modelling, make passably pretty pictures and display NMR structures.
Coot is Free Software. You can give it away. If you don't like the way it behaves, you can fix it yourself.
Coot is not:
The code is designed to be portable to any Unix-like operating system. Coot certainly runs on SGI IRIX64, RedHat Linux of various sorts, SuSe Linux4 and MacOS X (10.2). The sgi Coot binaries should also work on IRIX.
If you want to port to some other operating system, you are welcome 5. Note that your task will be eased by using GNU GCC to compile the programs components.
Coot works best with a 3-button mouse and works better if it has a scroll-wheel too (see Chapter 2 for more details) 6.
Coot responds to several environment variables that modify its behaviour.
And of course extension language environment variables are used too:
Normally, these environment variables will be set correctly in the coot shell script.
Rather that using the GUI to read in information, you can use the following command line arguments:
--c cmd to run a command cmd on start up
--script filename to run a script on start up (but see Section Scripting)
--no-state-script don't run the 0-coot.state.scm script on start up.
Don't save a state script on exit either.
--pdb filename for pdb/coordinates file
--coords filename for SHELX .ins/.res and CIF files
--data filename for mtz, phs or mmCIF data file
--auto filename for auto-reading mtz files (mtz file has the default labels FWT, PHWT)
--map filename for a map (currently CCP4-format only)
--dictionary filename read in a cif monomer dictionary
--help print command line options
--stereo start up in hardware stereo mode
--version print the version of coot and exit
--code accession-code on starting Coot, get the pdb file
and mtz file (if it exists) from the EDS
--no-guano don't leave “Coot droppings” i.e. don't write state
and history files on exit
--side-by-side start in side-by-side stereo mode
--update-self command-line mode to update the coot to the latest
pre-release on the server
--python an argument with no parameters - used to tell Coot that
the -c arguments should be process as python (rather than as scheme).
--small-screen start with smaller icons and font to fit on small
screen displays
--zalman-stereo start in Zalman stereo mode
So, for example, one might use:
coot --pdb post-refinement.pdb --auto refmac-2.mtz --dictionary lig.cif
There you can read more about the CCP4 molecular graphics project in general and other projects which are important for Coot 8.
Coot might crash on you - it shouldn't.
Whenever Coot manipulates the model, it saves a backup pdb file. There are backup files in the directory coot-backup 9. You can recover the session (until the last edit) by reading in the pdb file that you started with last time and then use File -> Recover Session....
I would like to know about coot crashing 10 so that I can fix it as soon as possible. If you want your problem fixed, this involves some work on your part sadly.
First please make sure that you are using the most recent version of coot. I will often need to know as much as possible about what you did to cause the bug. If you can reproduce the bug and send me the files that are needed to cause it, I can almost certainly fix it 11 - especially if you use the debugger (gdb) and send a backtrace too12. Note that you may have to source the contents of bin/coot so that the libraries are can be found when the executable dynamically links.
How do we move around and select things?
See also Chapter Hints and Usage Tips for more help.
See also “Recentring View” (Section Recentring View).
Use <+> or <-> on the keyboard if you don't have a scroll-wheel.
Here we can change the clipping and Translate in Screen Z
When there is no map, using the scroll-wheel has no effect. If there is exactly one map displayed, the scroll-wheel will change the contour level of that map. If there are two or more maps, the map for which the contour level is changed can be set using either HID -> Scrollwheel -> Attach scroll-wheel to which map? and selecting a map number or clicking the "Scroll" radio button for the map in the Display Manager.
You can turn off the map contour level changing by the scroll wheel using:
(set-scroll-by-wheel-mouse 0)
(the default is 1 [on]).
Several Coot functions require the selecting of atoms to specify a residue range (for example: Regularize, Refine (Section Regularization and Real Space Refinement) or Rigid Body Fit Zone (Section Rigid Body Refinement)). Select atoms with the Left-mouse. See also Picking (Section sec_picking).
Use the scripting function
(quanta-buttons) to make the mouse
functions more like other molecular graphics programs to which you may
be more accustomed 13.
You may not completely like the way the
molecule is moved by the mouse movement 14. To change this, try:
HID -> Virtual Trackball -> Flat. To
do this from the scripting interface: (vt-surface
1) 15.
If you do want screen-z rotation screen-z rotation, you can either use Shift Right-Mouse Drag or set the Virtual Trackball to Spherical Surface mode and move the mouse along the bottom edge of the screen.
The function (quanta-like-zoom) adds the ability to zoom the
view using just Shift + Mouse movement 16.
There is also a Zoom slider (Draw -> Zoom) for those without a right-mouse button.
The map-fitting and model-building tools can be accessed by using Calculate -> Model/Fit/Refine.... Many functions have tooltips 17 describing the particular features and are documented in Chapter Modelling and Building.
The version number of Coot can be found at the top of the “About” window (Help -> About).
This will return the version of coot:
$ coot --version
There is also a script function to return the version of coot:
(coot-version)
The built-in antialiasing (for what it's worth) can be enabled using:
(set-do-anti-aliasing 1)
The default is 0 (off).
This can also be activated using Edit Preferences -> Others -> Antialiasing -> Yes.
If you have an nVidia graphics card, external antialiasing can be actived setting the environment variable __GL_FSAA_MODE. For me a setting of 5 works nicely and gives a better image than using Coot's built-in antialiasing.
Also for nVidia graphics card users, there is the application nvidia-settings:
Antialiasing Setting -> Override Application Settings and slide the slider to the right. On restarting Coot, it should be in antialias mode 18.
Coot is based on the concept of molecules. Maps and coordinates are different representations of molecules. The access to the molecule is via the molecule number. It is often important therefore to know the molecule number of a particular molecule.
The Molecule Number of a molecule can be found by clicking on an atom of that molecule (if it has coordinates of course). The first number in brackets in the resulting text in the status bar and console is the Molecule Number. The Molecule Number can also be found in Display Control window (Section Display Manager). It is also displayed on the left-hand side of the molecule name in the option menus of the “Save Coordinates” and “Go To Atom” windows.
The “graphics” window is drawn using OpenGL. It is considerably smoother (i.e. more frames/sec) when using a 3D accelerated X server.
The view is orthographic (i.e. the back is the same size as the front). The default clipping is about right for viewing coordinate data, but is often a little too “thick” for viewing electron density. It is easily changed (see Section Clipping Manipulation).
Depth-cueing is linear and fixed on.
The graphics window can be resized, but it has a minimum size of 400x400 pixels.
Hardware Stereo is an option for Coot (Draw -> Stereo... -> Hardware Stereo -> OK), side-by-side stereo is not an option.
The angle between the stereo pairs (the stereo separation) can be changed to suit your personal tastes using:
(set-hardware-stereo-angle-factor angle-factor)
where angle-factor would typically be between 1.0 and 2.0
When asked to pick a residue or atom, the cursor changes from the normal arrow shape to a "pick" cursor. Sometimes it is difficult to see the default pick cursor, so you can change it using the function
(set-pick-cursor-index i)
where i is an integer less than 256. The cursors can be viewed using an external X program:
xfd -fn cursor
A yellow box called the “origin marker” marks the origin. It can be removed using:
(set-show-origin-marker 0)
Its state can be queried like this:
(show-origin-marker-state)
which returns an number (0 if it is not displayed, 1 if it is).
A simple screenshot (image dump) can be made using Draw -> Screenshot -> Simple.... Note that in side by side stereo mode you only get the left-hand image.
Output suitable for use by Raster3D's “render” can be generated using the scripting function
(raster3d file-name)
where file-name is such as "test.r3d"
19.
There is a keyboard key to generate this file, run “render” and display the image: Function key F8.
You can also use the function
(render-image)
which will create a file coot.r3d, from which “render” produces coot.png. This png file is displayed using ImageMagick's display program (by default). Use something like:
(set! coot-png-display-program "gqview")
to change that to different display program ("gqview" in this case).
(set! coot-png-display-program "open")
would use Preview (by default) on Macintosh.
To change the widths of the bonds and density “lines” use (for example):
(set-raster3d-bond-thickness 0.1)
and
(set-raster3d-density-thickness 0.01)
Similarly for bones:
(set-raster3d-bone-thickness 0.05)
To turn off the representations of the atoms (spheres):
(set-renderer-show-atoms 0)
This is also known as “Map and molecule (coordinates) display control”. Here you can select which maps and molecules you can see and how they are drawn 20. The “Display” and “Active” are toggle buttons, either depressed (active) or undepressed (inactive). The “Display” buttons control whether a molecule (or map) is drawn and the “Active” button controls if the molecule is clickable 21 (i.e. if the molecule's atoms can be labeled).
The "Scroll" radio buttons sets which map is has its contour level changed by scrolling the mouse scroll wheel.
By default, the path names of the files are not displayed in the Display Manager. To turn them on:
(set-show-paths-in-display-manager 1)
If you pull across the horizontal scrollbar in a Molecule view, you will see the “Render as” menu. You can use this to change between normal “Bonds (Colour by Atom)”,“Bonds (Colour by Chain)” and “C\alpha” representation There is also available “No Waters” and “C\alpha + ligands” representations.
You might not want to have the right-hand-side vertical toolbar that contains icons for some modelling operations 22 displayed:
(hide-modelling-toolbar)
to bring it back again:
(show-modelling-toolbar)
The “Filter” button in the fileselection filters the filenames according to extension. For coordinates files the extensions are “.pdb” “.brk” “.mmcif” and others. For data: “.mtz”, “.hkl”, “.phs”, “.cif” and for (CCP4) maps “.ext”, “.msk” and “.map”. If you want to add to the extensions, the following functions are available:
(add-coordinates-glob-extension extension)
(add-data-glob-extension extension)
(add-map-glob-extension extension)
(add-dictionary-glob-extension extension)
".mycif".
If you want the fileselection to be filtered without having to use the "Filter" button, use the scripting function
(set-filter-fileselection-filenames 1)
If you like your files initially sorted by date (rather than lexicographically, which is the default) use:
(set-sticky-sort-by-date)
Some people prefer that the fileselection for saving coordinates starts in the original directory (rather than the directory from which they last imported coordinates). This option is for them:
(set-save-coordinates-in-original-directory 1)
There is an compile-time option of adding a script interpreter. Currently the options are python and guile. It seems possible that in future you will be able to use both in the same executable. The binary distribution of Coot are linked with guile, others with python.
Hundreds of commands are made available for use in scripting by using SWIG, some of which are documented here. Other functions documented less well, but descriptions for them can be found at the end of this manual.
Commands described throughout this manual (such as (vt-surface
1)) can be evaluated
directly by Coot by using the “Scripting Window” (Calculate
-> Scripting...). Note that you type the commands in the upper
entry widget and the command gets echoed (in red) and the return value
and any output is displayed in the text widget lower (green). The typed
command should be terminated with a carriage return 23. Files 24 can be evaluated (executed)
using Calculate -> Run Script....
Note that in scheme (the usual scripting language of Coot), the parentheses are important.
To execute a script file from the command line use the --script
filename arguments
(except when also using the command line
argument --no-graphics, in which case you should use -s
filename).
After you have used the scripting window, you may have noticed that you can no longer kill Coot by using Ctrl-C in the console. To recover this ability:
(exit)
in the scripting window.
* Python Commands
Coot has an (optional) embedded python interpreter.
Thus the full power of python is available to you. Coot will look for
an initialization script
($HOME/.coot.py) and
will execute it if found. This file should contain python commands
that set your personal preferences.
The scripting functions described in this manual are formatted suitable for use with guile, i.e.:
(function arg1 arg2...)
If you are using Python instead: the format needs to be changed to:
function(arg1,arg2...)
Note that dashes in guile function names become underscores for
python, so that (for example) (raster-screen-shot) becomes
raster_screen_shot().
The scheme interpreter is made available by embedding
guile. The initialization script used by this interpreter is
$HOME/.coot. This file should contain scheme commands that
set your personal preferences.
The “state”
of Coot is saved on Exit and written to a
file called 0-coot.state.scm (scheme)
0-coot.state.py (python). This
state file contains information about the screen centre, the
clipping, colour map rotation size, the symmetry radius, and other
molecule related parameters such as filename, column labels,
coordinate filename etc..
Use Calculate -> Run Script... to use this file
to re-create the loaded maps and models that you had when you finished
using Coot 25 last time.
A state file can be saved at any time using (save-state)
which saves to file 0-coot.state.scm or
(save-state-filename "thing.scm") which saves to file
thing.scm.
When Coot starts it can optionally run the commands in
0-coot.state.scm.
Use (set-run-state-file-status i)
to change the behaviour: i is 0 to never run this
state file at
startup, i is
1 to get a dialog option (this is the default) and i
is 2 to run the commands without question.
“Power users” of Coot might like to write their own functions and bind that function to a keyboard key. How do they do that?
By using the add-key-binding function:
(add-key-binding function-name key function)
where key is a quoted string (note that upper case and lower case keys are distinguished - activate get upper case key binding you need to chord the shift key 26).
for example:
(add-key-binding "Refine Active Residue with Auto-accept" "x" refine-active-residue)
Have a look at the key bindings section on the Coot wiki for several more examples.
“Power users” of Coot might also like to write their own functions that occur after picking an atom (or a number of atoms)
(user-defined-click n_clicks udfunc)
define a function func which runs after the user has made n_clicked atom picks. func is called with a list of atom specifiers - the first member of which is the molecule number.
* Redo:: * Restoring from Backup::
By default, each time a modification is made to a model, the old coordinates are written out 27. The backups are kept in a backup directory and are tagged with the date and the history number (lower numbers are more ancient 28). The “Undo” function discards the current molecule and loads itself from the most recent backup coordinates. Thus you do not have to remember to “Save Changes” - coot will do it for you 29.
If you have made changes to more than one molecule, Coot will pop-up a dialog box in which you should set the “Undo Molecule” i.e. the molecule to which the Undo operations will apply. Further Undo operations will continue to apply to this molecule until there are none left. If another Undo is requested Coot checks to see if there are other molecules that can be undone, if there is exactly one, then that molecule becomes the “Undo Molecule”, if there are more than one, then another Undo selection dialog will be displayed.
You can set the undo molecule using the scripting function:
(set-undo-molecule imol)
If for reasons of strange system30 requirements you want to remove the path components of the backup file name you can do so using:
(set-unpathed-backup-file-names 1)
The “undone” modifications can be re-done using this button. This is not available immediately after a modification 31.
There may be certain
circumstances 32 in which you
wish to restore from a backup but can't get it by the “Undo”
mechanism described above. In that case, start coot as normal and
then open the (typically most recent) coordinates file in the
directory coot-backup (or the directory pointed to the
environment variable COOT_BACKUP_DIR if it was set) .
This file should contain your most recent edits. In such a case, it
is sensible for neatness purposes to immediately save the coordinates
(probably to the current directory) so that you are not modifying a
file in the backup directory.
See also Section Crash.
It is sometimes useful to use this to orient the view and export this orientation to other programs. The orientation matrix of the view can be displayed (in the console) using:
(view-matrix)
Also, the internal representation of the view can be returned and set using:
(view-quaternion) to return a 4-element list
(set-view-quaternion i j k l) which sets the view quaternion.
So the usage of these functions would be something like:
(let ((v (view-quaternion))) ;; manipulate v here, maybe (apply set-view-quaternion v))
Occasionally you may want to know the space group of a particular molecule. Interactively (for maps) you can see it using the Map Properties button in the Molecule Display Control dialog.
There is a scripting interface function that returns the space group for a given molecule 33:
(show-spacegroup imol)
You can force a space group onto a molecule using the following:
(set-space-group imol space-group)
where space-group is one of the standard CCP4 space group names (e.g. "P 21 21 21").
To show the symmetry operators of a particular molecule use:
(get-symmetry imol)
which will return a list of strings.
(set-rotation-centre x y z).
If you don't want smooth recentring (sliding) Edit -> Preferences -> Smooth Recentring -> Off. You can also use this dialog to speed it up a bit (by decreasing the number of steps instead of turning it off).
Coot has a views interface (you might call them ”scenes“) that define a particular orientation, zoom and view centre. Coot and linearly interpolate between the views. The animation play back speed can be set with the ”Views Play Speed“ menu item - default is a speed of 10.
The views interface can be found under the Extensions menu item.
The clipping planes (a.k.a. “slab” ) can be adjusted using Edit -> Clipping and adjusting the slider. There is only one parameter to change and it affects both the front and the back clipping planes 34. The clipping can also be changed using keyboard “D” and “F”.
It can also be changed with Ctrl + Right-mouse drag up and down. Likewise the screen-Z can be changed with Ctrl + Right-mouse left and right 35.
One can “push” and “pull” the view in the screen-Z direction using keypad 3 and keypad “.” (see Section Keyboard Z Translation).
The background colour can be set either using a GUI dialog
(Edit$ -> Background Colour) or the function
(set-background-colour 0.00 0.00 0.00), where the arguments
are 3 numbers between 0.0 and 1.0, which respectively represent the
red, green and blue components of the background colour. The default
is (0.0, 0.0, 0.0) (black).
If coordinates have symmetry available then unit cells can be drawn for molecules (Draw -> Cell & Symmetry -> Show Unit Cell?).
There is a pink pointer
at the centre of the screen that marks the rotation centre.
The size of the pointer can be changed using Edit
-> Pink Pointer Size... or using scripting commands:
(set-rotation-centre-size 0.3).
The Rotation Centre Pointer is sometimes called simply “Pointer”. One can find distances to the pointer from any active set of atoms using “Pointer Distances” (under Measures). If you move the Pointer (e.g. by centering on an atom) and want to update the distances to it, you have to toggle off and on the “Show Pointer Distances” on the Pointer Distances dialog.
Crosshairs can be drawn at the centre of the screen, using either the <C> key36 in graphics window or Draw -> Crosshairs.... The ticks are at 1.54Å, 2.7Å and 3.8Å.
Positions in 3D space can be annotated with 3D text. The mechanism to do this can be found under Extensions -> Representations -> 3D Annotations. 3D Annotations can be saved to and loaded from a file.
Sometimes, you might as yourself “how fast is the computer?”
37. Using Calculate ->
Frames/Sec you can see how fast the molecule is rotating, giving an
indication of graphics performance. It is often better to use a map
that is more realistic and stop the picture whizzing round. The output
is written to the status bar and the console, you need to give it a few
seconds to “settle down”. It is best not to have other widgets
overlaying the GL canvas as you do this.
The contouring elapsed time 38 gives an indication of CPU performance.
Due to its “in development” nature (at the moment), Coot produces a lot of “console” 39 output - much of it debugging or “informational”. This will go away in due course. You are advised to run Coot so that you can see the console and the graphics window at the same time, since feedback from atom clicking (for example) is often written there rather than displayed in the graphics window.
The format of coordinates that can be read by coot is either PDB or mmCIF. To read coordinates, choose File -> Read Coordinates from the menu-bar. Immediately after the coordinates have been read, the view is (by default) recentred to the centre of this new molecule and the molecule is displayed. The recentring of the view after the coordinates have been read can be turned off by unclicking the "Recentre?" radio-button.
To disable the recentring of the view on reading a coordinates file via
scripting, use: (set-recentre-on-read-pdb 0). However, when
reading a coordinates file from a script it is just as good (if not
better) to use (handle-read-draw-molecule-with-recentre
filename 0) - the additional 0 means “don't recentre”.
And that affects just the reading of filename and not
subsequent files.
Note that as of version 0.6.2 Coot can read MDL mol/mol2 files (the atom names are not unique (of course), but at least you can see the coordinates).
Coot uses the space group on the “CRYST1” line of the pdb file. The space group should be one of the xHM symbols listed (for example) in the CCP4 dictionary file syminfo.lib. So, for example, "R 3 2 :H" should be used in preference to "H32".
The reading multiple files using the GUI is not available (at the moment). However the following scripting functions are available:
(read-pdb-all)
which reads all the “*.pdb” files in the current directory
(multi-read-pdb glob-pattern dir)
which reads all the files matching glob-pattern in directory dir. Typical usage of this might be:
(multi-read-pdb "a*.pdb" ".")
Alternatively you can specify the files to be opened on the command line when you start coot (see Section Command Line Arguments).
SHELX ".res" (and ".ins" of course) files can be read into Coot, either using the GUI File -> Open Coordinates... or by the scripting function:
(read-shelx-ins-file file-name)
where file-name is quoted, such as "thox.ins".
Although Coot should be able to read any SHELX ".res" file, it may currently have trouble displaying the bonds for centro-symmetric structures.
ShelxL atoms with negative PART numbers are given alternative configuration identifiers in lower case.
To write a SHELX ".ins" file:
(write-shelx-ins-file imol file-name)
where imol is the number of the molecule you wish to export.
This will be a rudimentary file if the coordinates were initially from a "PDB" file, but will contain substantial SHELX commands if the coordinates were initially generated from a SHELX ins file.
Information about about a particular atom is displayed in the text console when you click using middle-mouse. Information for all the atoms in a residue is available using Info -> Residue Info....
The temperature factors and occupancy of the atoms in a residue can be set by using Edit -> Residue Info....
Use Shift + left-mouse to label atom. Do the same to toggle off the label. The font size is changeable using Edit -> Font Size.... The newly centred atom is labelled by default. To turn this off use:
(set-label-on-recentre-flag 0)
Some people prefer to have atom labels that are shorter, without the slashes and residue name:
(set-brief-atom-labels 1)
To change the atom label colour, use:
(set-font-colour 0.9 0.9 0.9)
The atom colouring system in coot is unsophisticated. Typically, atoms are coloured by element: carbons are yellow, oxygens red, nitrogens blue, hydrogens white and everything else green (see Section Display Manager for colour by chain). However, it is useful to be able to distinguish different molecules by colour, so by default coot rotates the colour map of the atoms (i.e. changes the H value in the HSV 40 colour system). The amount of the rotation depends on the molecule number and a user-settable parameter:
(set-colour-map-rotation-on-read-pdb 30).
The default value is 31^\circ.
Also one is able to select only the Carbon atoms to change colour in
this manner: (set-colour-map-rotation-on-read-pdb-c-only-flag
1).
The colour map rotation can be set individually for each molecule by
using the GUI: Edit -> Bond Colours....
The various bond parameters can be set using the GUI dialog Draw -> Bond Parameters or via scripting functions.
The represention style of the molecule that has the active residue (if any) can be changed using the scroll wheel with Ctrl and Shift.
The thickness (width) of bonds of individual molecules can be changed. This can be done via the Bond Parameters dialog or the scripting interface:
(set-bond-thickness thickness imol)
where imol is the molecule number.
The default thickness is 3 pixels. The bond thickness also applies to the symmetry atoms of the molecule. The default bond thickness for new molecules can be set using:
(set-default-bond-thickness thick)
where thick is an integer.
There is no means to change the bond thickness of a residue selection within a molecule.
Initially, hydrogens are displayed. They can be undisplayed using
(set-draw-hydrogens mol-no 0) 41
where mol-no is the molecule number.
There is a GUI to control this too, under “Edit -> Bond Parameters”.
It is occasionally useful when analysing non-crystallographically related molecules to have “images” of the other related molecules appear matched onto the current coordinates. It is important to understand that these ghosts are for displaying differences of NCS-related molecules by structure superposition, not displaying neighbouring NCS related molecules. As you read in coordinates in Coot, they are checked for NCS relationships and clicking on “Edit -> Bond Parameters -> Show NCS Ghosts” -> “Yes” -> “Apply” will create “ghost” copies of them over the reference chain 42.
Sometimes SSM does not provide a good (or even useful) matrix. In that case, we can specify the residue range ourselves and let the LSQ algorithm provide the matrix. A gui dialog for this operation can be found under Extensions -> NCS -> NCS Ghosts by Residue Range....
The scripting function is used like this:
(manual-ncs-ghosts imol resno-start resno-end ncs-chain-ids)
Typical usage: (manual-ncs-ghosts 0 1 10 (list "A" "B" "C"))
note that in ncs-chain-ids, the NCS master/reference chain-id goes first.
Coot can use the relative transformations of the NCS-related molecules in a coordinates molecule to transform maps. Use Calculate -> NCS Maps... to do this (note the NCS maps only make sense in the region of the reference chain (see above).
Note also that the internal representation of the map is not transformed. If you try to export a NCS overlay map you will get an untransformed map. A transformed map only makes sense around a given point (and when using transformed maps in Coot, this reference point is changed on the fly, thus allowing map transformations on the fly). [This applies to NCS overlap maps, NCS averaged maps are transformed].
This will also create an NCS averaged map 43.
Coot can use a set of strict NCS matrices to specify NCS which means that NCS-related molecules can appear like convention symmetry-related molecules.
(add-strict-ncs-matrix imol ncs-chain-id ncs-target-chain-id m11 m12 m13 m21 m22 m23 m31 m32 m33 t1 t2 t3)
where ncs-chain-id might be "B", "C" "D" (etc.) and ncs-target-chain-id is "A", i.e. the B, C, D molecules are NCS copies of the A chain.
for icosahedral symmetry the translation components t1, t2, t3 will be 0.
You need to turn on symmetry for molecule imol and set the displayed symmetry object type to "Display Near Chains".
Coot provides the possibility to download coordinates from an OCA 44. (e.g. EBI) server 45 (File -> Get PDB Using Code...). A pop-up entry box is displayed into which you can type a PDB accession code. Coot will then connect to the web server and transfer the file. Coot blocks as it does this (which is not ideal) but on a semi-decent internet connection, it's not too bad. The downloaded coordinates are saved into a directory called coot-download.
It is also possible to download mmCIF data and generate a map. This currently requires a properly formatted database structure factors mmCIF file 46.
Using this function we have the ability to download coordinates and view the map from structures in the Electron Density Server (EDS) at Uppsala University. This is a much more robust and faster way to see maps from deposited structures. This function can be found under the File menu item.
This feature was added with the assistance of Gerard Kleywegt. If you use the EDS, please cite GJ Kleywegt, MR Harris, JY Zou, TC Taylor, A Wählby & TA Jones (2004), "The Uppsala Electron-Density Server", Acta Cryst. D60, 2240-2249.
On selecting from the menus File -> Save Coordinates... you are first presented with a list of molecules which have coordinates. As well as the molecule number, there is the molecule name - very frequently the name of the file that was read in to generate the coordinates in coot initially. However, this is only a molecule name and should not be confused with the filename to which the coordinates are saved. The coordinates filename can be selected using the Select Filename... button.
If your filename ends in .cif, .mmcif or
.mmCIF then an mmCIF file will be written (not a “PDB”
file).
If for some reason, the pdb file that you read does not have a space group, or has the wrong space group, then you can set it using the following function:
(set-space-group imol symbol)
e.g.:
(set-space-group 0 "P 41 21 2")
By default anisotropic atom information is
not represented 47. To turn them on,
use Draw -> Anisotropic Atoms -> Show
Anisotropic Atoms? -> Yes, or the command:
(set-show-aniso 1).
You cannot currently display thermal ellipsoids 48 for isotropic atoms.
Coordinates symmetry is “dynamic”. Symmetry atoms can be labeled 49. Every time you recentre, the symmetry coordinates are updated. The information shown contains the atom information and the symmetry operation number and translations needed to generate the atom in that position.
By default symmetry atoms are not displayed.
If you want coot to display symmetry coordinates without having to use the gui, add to your ~/.coot the following:
(set-show-symmetry-master 1)
The symmetry can be represented as C\alphas. This along with representation of the molecule as C\alphas (Section Display Manager) allow the production of a packing diagram.
Sometimes (rarely) coot misses symmetry-related molecules that should be displayed. In that case you need to expand the shift search (the default is 1):
(set-symmetry-shift-search-size 2)
This is a hack, until the symmetry search algorithm is improved.
The protein is represented by one letter codes and coloured according to secondary structure. These one letter codes are active - if you click on them, they will change the centre of the graphics window - in much the same way as clicking on a residue in the Ramachandran plot.
The single letter code (of the imolth molecule) is written out to the console in FASTA format. Use can use this to cut and paste into other applications:
(print-sequence imol)
Environment distances are turned on using Info -> Environment Distances.... Contacts to other residues are shown and to symmetry-related atoms if symmetry is being displayed. The contacts are coloured by atom type 50.
The distance between atoms can be found using Info -> Distance 51. The result is displayed graphically, and written to the console.
Atoms of zero occupancy are marked with a grey spot. To turn off these markers, use:
(set-draw-zero-occ-markers 0)
Use an argument of 1 to turn them on.
You can draw dots round arbitrary atom selections
(dots imol atom-selection dot-density radius)
The function returns a handle.
e.g. put a sphere of dots around all atoms of the 0th molecule (it might be a set of heavy atom coordinates) at the default dot density and radius:
(dots 0 "/1" 1 1)
You can't change the colour of the dots.
There is no internal mechanism to change the radius according to atom type. With some cleverness you might be able to call this function several times and change the radius according to the atom selection.
There is a function to clear up the dots for a particular molecule imol and dots set identifier dots-handle
(clear-dots imol dots-handle)
There is a function to return how many dots sets there are for a particular molecule imol:
(n-dots-set imol)
Fragments of the molecule can be rendered as a “ball and stick” molecule:
(make-ball-and-stick imol atom-selection bond-thickness sphere-size draw-spheres-flag)
e.g.
(make-ball-and-stick 0 "/1/A/10-20" 0.3 0.4 1)
The ball-and-stick representation can be cleared using:
(clear-ball-and-stick imol)
Coot can be used to calculate the mean (average) and median temperatures factors:
(average-temperature-factor imol)
(median-temperature-factor imol)
-1 is returned if there was a problem 52.
The excellent SSM alogrithm53 of Eugene Krissinel is available in Coot. The GUI interface is straight-forward and can be found under Calculate -> SSM Superpose. You can specify the specific chains that you wish to match using the "Use Specific Chain" check-button.
There is a scripting level function which gives even finer control:
(superpose-with-atom-selection imol1 imol2
mmdb-atom-selection-string-1 mmdb-atom-selection-string-2
move-copy-flag )
the move-copy-flag should be 1 if you want to apply the transformation to a copy of imol2 (rather than imol2 itself). Otherwise, move-copy-flag should be 0.
mmdb atom selection strings (Coordinate-IDs) are explained in detail in the mmdb manual.
Briefly, the string should be formed in this manner:
/mdl/chn/seq(res).ic/atm[elm]:aloc
e.g. "/1/A/12-130/CA"
<p><a href="http://www.ebi.ac.uk/~keb/cldoc/object/cl_obj_surf.html#CoordinateID">The mmdb manual CoordinateID description</a>.</p>
There is a simple GUI for this Calculate -> LSQ Superpose...
The scripting interface to LSQ fitting is as follows:
(simple-lsq-match ref-start-resno ref-end-resno ref-chain-id imol-ref
mov-start-resno mov-end-resno mov-chain-id imol-mov
match-type)
where:
'CA,
'main, or 'all.
e.g.:
(simple-lsq-match 940 950 "A" 0 940 950 "A" 1 'main)
More sophisticated (match molecule number 1 chain “B” on to molecule number 0 chain “A”):
(define match1 (list 840 850 "A" 440 450 "B" 'all))
(define match2 (list 940 950 "A" 540 550 "B" 'main))
(clear-lsq-matches)
(set-match-element match1)
(set-match-element match2)
(lsq-match 0 1) ; match molecule number 1 onto molecule number 0.
(overlap-ligands imol-ligand imol-ref chain-id-ref resno-ref)
returns a rotation+translation operator which can be applied to other molecules (and maps). Here, imol-ligand is the molecule number of the ligand (which is presumed to be a a molecule on its own - Coot simply takes the first residue that it finds). imol-ref chain-id-ref resno-ref collectively describe the target position for the moving imol-ligand molecule.
The convenience function
(overlay-my-ligands imol-mov chain-id-mov resno-mov imol-ref chain-id-ref resno-ref)
wraps overlap-ligands.
The GUI for the function can be found under
Extensions -> Modelling -> Supperpose Ligands...
As well as the GUI option File -> Save Coordinates... there is a scripting options available:
(write-pdb-file imol pdb-file-name)
which writes the imolth coordinates molecule to filename.
To write a specific residue range:
(write-residue-range-to-pdb-file imol chain-id start-resno
endresno pdb-file-name)
The functions described in this chapter manipulate, extend or build molecules and can be found under Calculate -> Model/Fit/Refine.... When activated, the dialog "stays on top" of the main graphics window 54. Some people think that this is not always desirable, so this behaviour can be undone using:
(set-model-fit-refine-dialog-stays-on-top 0)
Coot will read the geometry restraints for refmac and use them in fragment (zone) idealization - this is called “Regularization”. The geometrical restraints are, by default, bonds, angles, planes and non-bonded contacts. You can additionally use torsion restraints by Calculate -> Model/Fit/Refine... -> Refine/Regularize Control -> Use Torsion Restraints. Truth to tell, this has not been successful in my hands (sadly).
“RS (Real Space) Refinement” (after Diamond, 1971 55) in Coot is the use of the map in addition to geometry terms to improve the positions of the atoms. Select “Regularize” from the “Model/Fit/Refine” dialog and click on 2 atoms to define the zone (you can of course click on the same atom twice if you only want to regularize one residue). Coot then regularizes the residue range. At the end Coot, displays the intermediate atoms in white and also displays a dialog, in which you can accept or reject this regularization. In the console are displayed the \chi^2 values of the various geometrical restraints for the zone before and after the regularization. Usually the \chi^2 values are considerably decreased - structure idealization such as this should drive the \chi^2 values toward zero.
The use of “Refinement” is similar - with the addition of using a map. The map used to refine the structure is set by using the “Refine/Regularize Control” dialog. If you have read/created only one map into Coot, then that map will be used (there is no need to set it explicitly).
Use, for example,
(set-matrix 20.0)
to change the weight of the map gradients to geometric gradients. The higher the number the more weight that is given to the map terms 56. The default is 60.0. This will be needed for maps generated from data not on (or close to) the absolute scale or maps that have been scaled (for example so that the sigma level has been scaled to 1.0).
For both “Regularize Zone” and “Refine Zone” one is able to use a
single click to
refine a residue range. Pressing <A> on the keyboard while
selecting an atom in a residue will automatically create a residue
range with that residue in the middle. By default the zone is
extended one residue either size of the central residue. This can be
changed to 2 either side using (set-refine-auto-range-step
2).
Intermediate (white) atoms can be moved around with the mouse (click and drag with left-mouse, by default). Refinement will proceed from the new atom positions when the mouse button is released. It is possible to create incorrect atom nomenclature and/or chiral volumes in this manner - so some care must be taken. Press the <A> key as you left-mouse click to move atoms more “locally” (rather than a linear shear) and <Ctrl> key as you left-mouse click to move just one atom.
In more up to date versions, Coot will display colour patches (something like a traffic light system) representing the chi squared values of each of types of geometric feature refined. Typically “5 greens” is the thing to aim for, the colour changes occurring at chi squared values 2, 5 and 8 (8 being the most red).
To prevent the unintentional refinement of a large number of residues,
there is a “heuristic fencepost” of 20 residues. A selection of
than 20 residues will not be regularized or refined. The limit can be
changed using the scripting function: e.g.
(set-refine-max-residues 30).
The geometry description for residues, monomers and links used by Coot are in the standard mmCIF format. Because this format alows multiple comp_ids (residue types) to be described within a cif loop, it is hard to tell when a dictionary entry needs to be overwritten when reading a new file. Therefore Coot makes this extra constraint: that the “chem_comp” loop should appear first in the comp list data item - if this is the case, then Coot can overwrite an old restraint table for a particular comp_id/residue-type when a new one is read.
By default, the geometry dictionary entries for only the standard residues are read in at the start 57. It may be that your particular ligand is not amongst these. To interactively add a dictionary entry use File -> Import CIF Dictionary. Alternatively, you can use the function:
(read-cif-dictionary filename)
and add this to your .coot file (this may be the preferred
method if you want to read the file on more than one occasion).
Note: the dictionary also provides the description of the ligand's torsions.
Sphere refinement selects residues within a certain distance of the residue at the centre of the screen and includes them for real space refinement. In this way, one can select residues that are not in a linear range. This technique is useful for refining disulfide bonds and glycosidic linkages.
To enable sphere refinement, Right-mouse in the vertical toolbutton menu, Manage buttons -> [Tick] Sphere Refine -> Apply. You will need a python-enabled Coot to do this.
The following adds a key binding (Shift-R) that refines resides that are within 3.5Å of the residue at the centre of the screen:
(define *sphere-refine-radius* 3.5)
(add-key-binding "Refine residues in a sphere" "R"
(lambda ()
(using-active-atom
(let* ((rc-spec (list aa-chain-id aa-res-no aa-ins-code))
(ls (residues-near-residue aa-imol rc-spec *sphere-refine-radius*)))
(refine-residues aa-imol (cons rc-spec ls))))))
Refining carbohydrates monomers should be as straightforward as refining a protein residue. Coot will look in the dictionary for the 3-letter code for the particular residue type, if it does not find it, Coot will try to search for dictionary files using “-b-D” or “-a-L” extensions.
When refining a group of carbohydrates, the situation needs a bit more explanation. For each residue pair with tandem residue numbers specified in the refinement range selection, Coot checks if these residue types are are furanose or pyranose in the dictionary, and if the are both one or the other, then it tries to see if there are any of the 11 link types (BETA1-4, BETA2-3, ALPHA1-2 and so on) specified in the dictionary. It does this by a distance check of the potentially bonding atoms. If the distance is less than 3.0Å, then a glycosidic bond is made and used in the refinement.
Bonds between protein and carbohydrate and branched carbohydrates can be refined using “Sphere Refinement”.
LINK and LNKR cards are not used to determine the geometry of the restraints.
By default, Coot uses a 5 atom (CA-1, C-1, O-1, N-2, CA-2) planar peptide restraints. These restraints should help in low resolution fitting (the main-chains becomes less distorted), reduce accidental cis-peptides and may help “clean up” Ramachandran plots.
(add-planar-peptide-restraints)
And similarly they can be removed:
(remove-planar-peptide-restraints)
There is also a GUI to add and remove these restraints in
Extensions -> Refine... -> Peptide Restraints...
The UNK residue type is a special residue type to Coot. It has
been added for use with Buccaneer. Don't give you ligand (or anything
else) the 3-letter-code UNK or confusion will result
58.
By default, atoms with zero occupancy are moved when refining and regularizing. This can sometimes be inconvenient. To turn of the movement of atoms with zero occupancy when refining and regularizing:
(set-refinement-move-atoms-with-zero-occupancy 0)
You can change the map that is used for the fitting and refinement tools using the Select Map... button on the Model/Fit/Refine dialog.
“Rotate/Translate Zone” from the “Model/Fit/Refine” menu allows manual movement of a zone. After pressing the “Rotate/Translate Zone” button, select two atoms in the graphics canvas to define a residue range 59, the second atom that you click will be the local rotation centre for the zone. The atoms selected in the moving fragment have the same alternate conformation code as the first atom you click. To actuate a transformation, click and drag horizontally across the relevant button in the newly-created “Rotation & Translation” dialog. The axis system of the rotations and translations are the screen coordinates. Alternatively 60, you can click using left-mouse on an atom in the fragment and drag the fragment around. Use Control Left-mouse to move just one atom, rather than the whole fragment. If you click Control Left-mouse whilst not over an atom then you can rotate the fragment using mouse drag. Click “OK” (or press Return) when the transformation is complete.
To change the rotation point to the centre of the intermediate atoms (rather than the second clicked atom), use the setting:
(set-rotate-translate-zone-rotates-about-zone-centre 1)
“Rigid Body Fit Zone” from the “Model/Fit/Refine” dialog provides rigid body refinement. The selection is zone-based 61. So to refine just one residue, click on one atom twice.
Sometimes no results are displayed after Rigid Body Fit Zone. This is because the final model positions had too many final atom positions in negative density. If you want to over-rule the default fraction of atoms in the zone that have an acceptable fit (0.75), to be (say) 0.25:
(set-rigid-body-fit-acceptable-fit-fraction 0.25)
Rigid body refinement via Nelder-Mead Simplex minimization is available in Coot. Simplex refinement has a larger radius of convergence and thus is useful in a position where simple rigid body refinement finds the wrong minimum. However the Simplex algorithm is much slower. Simplex refinement for a residue range start-resno to end-resno (inclusive) in chain chain-id can be accessed as follows:
(fit-residue-range-to-map-by-simplex start-resno end-resno alt-loc
chain-id imol imol-for-map)
There is currently no GUI interface to Simplex refinement.
If you wanted automatically run a function after a model has been manipulated then you can do so using by creating a function that takes 2 arguments, such as:
(post-manipulation-hook imol manipulation-mode)
manipulation-mode is one of (DELETED), (MUTATED) or (MOVINGATOMS).
And of course imol is the model number of the maniplated molecule.
(It would of course be far more useful if this function was also passed a list of residues - that is something for the future).
Baton build is most useful if a skeleton is already calculated and displayed (see Section Skeletonization). When three or more atoms have been built in a chain, Coot will use a prior probability distribution for the next position based on the position of the previous three. The analysis is similar to that of Oldfield & Hubbard (1994) 62, however it is based on a more recent and considerably larger database.
Little crosses are drawn representing directions in which is is possible that the chain goes, and a baton is drawn from the current point to one of these new positions. If you don't like this particular direction 63, use Try Another. The list of directions is scored according to the above criterion and sorted so that the most likely is at the top of the list and displayed first as the baton direction.
When starting baton building, be sure to be about 3.8Å from the position of the first-placed C\alpha, this is because the next C\alpha is placed at the end of the baton, the baton root being at the centre of the screen. So, when trying to baton-build a chain starting at residue 1, centre the screen at about the position of residue 2.
It seems like a good idea to increase the map sampling to 2 or even 2.5 (before reading in your mtz file) [a grid sampling of about 0.5Å seems reasonable] when trying to baton-build a low resolution map. You can set the map sampling using Edit -> Map Parameters -> Map Sampling.
Occasionally, every point is not where you want to position the next atom. In that case you can either shorten or lengthen the baton, or position it yourself using the mouse. Use “b” on the keyboard to swap to baton mode for the mouse 64.
Baton-built atoms are placed into a molecule called “Baton Atom” and it is often sensible to save the coordinates of this molecule before quitting coot.
If you try to trace a high resolution map (1.5Å or better) you will need to increase the skeleton search depth from the default (10), for example:
(set-max-skeleton-search-depth 20)
Alternatively, you could generate a new map using data to a more moderate resolution (2Å), the map may be easier to interpret at that resolution anyhow 65.
The guide positions are updated every time the “Accept” button is clicked. The molecule name for these atoms is “Baton Build Guide Points” and is is not usually necessary to keep them.
There is also an “Undo” button for baton-building. Pressing this will delete the most recently placed C\alpha and the guide points will be recalculated for the previous position. The number of “Undo”s is unlimited. Note that you should use the “Undo” button in the Baton Build dialog, not the one in the “Model/Fit/Refine” dialog (Section Backups and Undo).
Sometimes (especially at loops) you can see the direction in which the chain should go, but there is no skeleton (see Section Skeletonization) is displayed (and consequently no guide points) in that direction. In that case, “Undo” the previous atom and decrease the skeletonization level (Edit -> Skeleton Parameters -> Skeletonization Level). Accept the atom (in the same place as last time) and now when the new guide points are displayed, there should be an option to build in a new direction.
The following scenario is not uncommon: you find a nice stretch of density and start baton building in it. After a while you come to a point where you stop (dismissing the baton build dialog). You want to go back to where you started and build the other way. How do you do that?
(set-baton-build-params start-resno
chain-id "backwards")
where start-resno would typically
be 0 66 and chain-id would be
"" (default).
After you've build a fragment, sometimes you might want to change the direction of that fragment (this function changes an already existing fragment, as opposed to Backwards Building which sets up Baton Building to place new points in reverse order).
The fragment is defined as a contiguous set of residues numbers. So that you should be sure that other partial fragments which have the same chain id and that are not connected to this fragment have residue numbers that are not contiguous with the fragment you are trying to reverse.
Mainchain can be generated using a set of C\alphas as guide-points (such as those from Baton-building) along the line of Esnouf 67 or Jones and coworkers 68. Briefly, 6-residue fragments of are generated from a list of high-quality 69 structures. The C\alpha atoms of these fragments are matched against overlapping sets of the guide-point C\alphas. The resulting matches are merged to provide positions for the mainchain (and C\beta) atoms. This procedure works well for helices and strands, but less well 70 for less common structural features.
This function is also available from the scripting interface:
(db-mainchain imol chain-id resno-start resno-end direction)
where direction is either "backwards" or "forwards".
Recall that the chain-id needs to be quoted, i.e.
use "A" not A. Note that chain-id is
"" when the C\alphas have been built with Baton Mode in
Coot.
It is possible to edit the backbone \phi and \psi angles indirectly using an option in the Model/Fit/Refine's dialog: “Edit Backbone Torsions..”. When clicked and an atom of a peptide is selected, this produces a new dialog that offers “Rotate Peptide” which changes this residues \psi and “Rotate Carbonyl” which changes \phi. Click and drag across the button 71 to rotate the moving atoms in the graphics window. You should know, of course, that making these modifications alter the \phi/\psi angles of more than one residue.
Docking sidechains means adding sidechains to a model or fragment that has currently only poly-Ala, where the sequence assignment is unknown. The algorithm is basically the same as in Cowtan's Buccaneer, but with some corners cut to make things (more or less) interactive. The algorithm uses the shape of the density around the C-beta position to estimate the probability of each sidechain type at that position.
The function is accessed via the Extensions -> Dock Sequence menu item. First, a sequence should be assigned from a PIR file to a particular chain-id and model number. Secondly Extensions -> Dock Sequence -> Dock Sequence on this fragment.... Choose the model to build on and then Dock Sequence! If all goes well, the model will be updated with mutated residues and undergo rotamer seach for each of the new residues. If the sequence alignment is not sufficiently clear, then you will get a dialog suggesting that you extend or improve the fragment.
The rotamers are generated 72 from the backbone independent sidechain library of the Richardsons group 73.
The m, t and p stand for “minus (-60)”, “trans (180)” and “plus (+60)”. There is one letter per \chi angle.
Use keyboard <.> and <,> to cycle round the rotamers.
“Auto Fit Rotamer” will try to fit the rotamer to the electron density. Each rotamer is generated, rigid body refined and scored according to the fit to the map. Fitting the second conformation of a dual conformation in this way will often fail - the algorithm will pick the best fit to the density - ignoring the position of the other atoms.
The algorithm doesn't know if the other atoms in the structure are in
sensible positions. If they are, then it is sensible not to put this
residue too close to them, if they are not then there should be no
restriction from the other atoms as to the position of this residue -
the default is “are sensible”, which means that the algorithm is
prevented from finding solutions that are too close to the atoms of
other residues. (set-rotamer-check-clashes 0) will stop this.
There is a scripting interface to auto-fitting rotamers:
(auto-fit-best-rotamer resno alt-loc ins-code chain-id
imol-coords
imol-map clash-flag lowest-rotamer-probability)
where:
resno is the residue number
alt-loc is the alternate/alternative location symbol
(e.g. "A" or "B", but most often "")
ins-code is the insertion code (usually "")
imol-coords is the molecule number of the coordinates molecule
imol-map is the molecule number of the map to which you wish to fit the side chains
clash-flag should the positions of other residues be included in the scoring of the rotamers (i.e. clashing with other other atoms gets marked as bad/unlikely)
lowest-rotamer-probability: some rotamers of some side chains are so unlikely that they shouldn't be considered - typically 0.01 (1%).
You can change the auto-fit rotamer fitting algorithms using
(set-rotamer-search-mode mode)
where mode is one of (ROTAMERSEARCHAUTOMATIC),
(ROTAMERSEARCHLOWRES) (i.e. "Backrub Rotamers"
(vide infra)) or (ROTAMERSEARCHHIGHRES) (the
conventional/high-resolution method using rigid-body fitting).
By default, the auto-fit rotamer method is (ROTAMERSEARCHAUTOMATIC).
By default, Auto Fit Rotamer will switch to “Backrub Rotamer” 74 mode when fitting against a map of worse than 2.7Å. This search mode moves the some atoms of the mainchain of the neighbouring residues. After rotation of the central residue and neighbouring atoms around the “backrub vector”, the individual peptides are back-rotated (along the peptide axis) so that the carbonyl oxygen are placed as near as possible to their original position. The Ramachandran plot is not used in this fitting algorithm.
Sometimes you don't have a map 75 but nevertheless there are clashing residues 76 (for example after mutation of a residue range) and you need to rotate side-chains to a non-clashing rotamer. There is a scripting interface:
(de-clash imol chain-id start-resno end-resno)
start-resno is the residue number of the first residue you wish to de-clash
end-resno is the residue number of the last residue you wish to de-clash
imol is the molecule number of the coordinates molecule
This interface will not change residues with insertion codes or alternate conformation. The lowest-rotamer-probability is set to 0.01.
Instead of using Rotamers, one can instead change the \chi angles (often called “torsions”) “by hand” (using “Edit Chi Angles” from the “Model/Fit/Refine” dialog). To edit a residue's \chi_1 press “1”: to edit \chi_2, “2”: \chi_3 “3” and \chi_4 “4”. Use left-mouse click and drag to change the \chi value. Use keyboard “0” 77 to go back to ordinary view mode at any time during the editing. Alternatively, one can use the “View Rotation Mode” or use the <Ctrl> key when moving the mouse in the graphics window. Use the Accept/Reject dialog when you have finished editing the \chi angles.
For non-standard residues, the clicked atom defines the base of the atom tree, which defines the “head” of the molecule (it's the “tail” (twigs/leaves) that wags). To emphasise, then: it matters on which atom you click!
By default torsions for hydrogen atoms are turned off. To turn them on:
(set-find-hydrogen-torsions 1)
To edit the rotatable bonds of a ligand using this tool, you will need to have read in the mmCIF dictionary beforehand.
You need to click on the torsion-general button, then click 4 atoms that describe the torsion - the first atom will be the base (non moving) part of the atom tree, on clicking the 4th atom a dialog will pop up with a "Reverse" button. Move this dialog out of the way and then left mouse click and drag in the main window will rotate the "top" part of the residue round the clicked atoms 2 and 3. When you are happy, click "Accept".
If you are torsion generaling a residue that has an alt conf, then the atoms of residue that are moved are those that have the same alt conf as the 4th clicked atom (or have an blank alt conf).
For ligands, you will need to read the mmCIF file that contains a description of the ligand's geometry (see Section Regularization and Real Space Refinement). By default, torsions that move hydrogens are not included. Only 9 torsion angles are available from the keyboard torsion angle selection.
Coot uses the same pepflip scheme
as is used in O (i.e. the C, N and O atoms are rotated
180^o round a line joining the C\alpha atoms of the residues
involved in the peptide). Flip the peptide again to return the atoms
to their previous position.
The allows the addition alternate (dual, triple etc.) conformations to the picked residue. By default, this provides a choice of rotamer (Section Rotamers). If there are not the correct main chain atoms a rotamer choice cannot be provided, and Coot falls back to providing intermediate atoms.
The default occupancy for new atoms is 0.5. This can be changed by using use slider on the rotamer selection window or by using the scripting function:
(set-add-alt-conf-new-atoms-occupancy 0.4)
The remaining occupancy of the atoms (after the new occupancy has been added) is split amongst the atoms that existed in the residue before the split. It is important therefore that the residues atoms have sane occupancies before adding an alternative conformation.
The default Split Type is to split the whole residue. If you want the default to be to split a residue after (and including) the CA, then add to your .coot file:
(set-add-alt-conf-split-type-number 0)
Mutations are available on a 1-by-1 basis using the graphics. After selecting “Mutate...” from the “Model/Fit/Refine” dialog, click on an atom in the graphics. A “Residue Type” window will now appear. Select the new residue type you wish and the residue in the graphics is updated to the new residue type 78. The initial position of the new rotamer is the a priori most likely rotamer. Note that in interactive mode, such as this, a residue type match 79 will not stop the mutation action occurring.
Mutation of DNA or RNA can be performed using “Simple Mutate” from the Model/Fit/Refine dialog. Residues need to be named "Ad", "Gr", "Ur" etc.
This dialog can be found under Calculate -> Mutate Residue Range. A residue range can be assigned a sequence and optionally fitted to the map. This is useful converting a poly-ALA model to the correct sequence 80.
Multiple mutations
are also supported via
the scripting interface. Unlike the single residue mutation function,
a residue type match will prevent a modification of the
residue 81.
Two functions are provided: To mutate a whole chain, use
(mutate-chain imol chain-id sequence) where:
chain-id is the chain identifier of the chain that you wish
to mutate (e.g. "A") and
imol is molecule number.
sequence is a list of single-letter residue codes,
such as "GYRESDF" (this should be a straight string with no
additional spaces or carriage returns).
Note that the number of residues in the sequence chain and those in the chain of the protein must match exactly (i.e. the whole of the chain is mutated (except residues that have a matching residue type).)
To mutate a residue range, use
(mutate-residue-range imol chain-id
start-res-no stop-res-no sequence)
where
start-res-no is the starting residue for mutation
stop-res-no is the last residue for mutation, i.e. using values of 2 and 3 for start-res-no and stop-res-no respectively will mutate 2 residues.
Again, the length of the sequence must correspond to the residue range length. Note also that this is a protein sequence - not nucleic acid.
For mutation of nucleic acids, use:
(mutate-nucleotide-range imol chain-id resno-start
resno-end sequence)
Sometimes one might like to model post-translational or other such modifications. How is that done, if the new residue type is not one of the standard residue types?
There is a scripting function:
(mutate-by-overlap imol chain-id resno new-three-letter-code)
This imports a model residue for the new residue type and overlays it on to the given residue by using graph-matching to determine the equivalent atoms.
The GUI for this can be found under Extensions -> Modelling -> Replace Residue... (for this to work, you need to be centred on the residue you wish to replace).
Note that if you are replacing are conventional protein residue with a
modified form (e.g. replacing a TYR with a phoso-tyrosine or a
LYS with an acetyl-lysine) you will need to make sure that the group
of the resulting restraints is an L-peptide (use Edit
-> Restraints to check and modify the restraints group. Likewise for
modified RNA/DNA nucleotides, you need to specify the group as
RNA or DNA as appropriate.
The function combines Mutation and Auto Fit Rotamer and is the easiest way to make a mutation and then fit to the map. You can currently only “Mutate and Autofit” protein residues (i.e. things with a rotamer dictionary.
Renumbering is straightforward using the renumber dialog available under Calculate -> Renumber Residue Range.... There is also a scripting interface:
(renumber-residue-range imol chain-id start-res-no
last-resno offset)
You can import monomers (often ligands) using File -> Get Monomer...82 by providing the 3-letter code of your monomer/ligand. The resulting molecule will be moved so that it placed at the current screen centre.
Typically, when you are happy about the placement of the ligand, you'd then use Merge Molecules to add the ligand/monomer to the main set of coordinates.
This procedure creates a pdb file monomer-XXX.pdb and a dictionary file libcheck_XXX.cif in the directory in which Coot was started.
A future invocation of Get Monomer uses these file so that the monomer appears quickly 83.
Similarly, you can generate ligands using File -> SMILES... and providing a SMILES string and a code for the residue name (this is your name for the residue type and a dictionary will be generated for the monomer of this type). This function is also a wrapper to LIBCHECK.
You are offered a selection of maps to search (you can only choose one at a time) and a selection of molecules that act as a mask to this map. Finally you must choose which ligand types you are going to search for in this map 84. Only molecules with less than 400 atoms are suggested as potential ligands.
If you do not have any molecules with less that 400 atoms loaded in Coot, you will get the message:
"Error: you must have at least one ligand to search for!"
New ligands are placed where the map density is and protein (mask) atoms are not). The masked map is searched for clusters using a default cut-off of 1.0\sigma. In weak density this cut-off may be too high and in such a case the cut-off value can be changed using something such as:
(set-ligand-cluster-sigma-level 0.8)
However, if the map to be searched for ligands is a difference map, a cluster level of 2.0 or 3.0 would probably be more appropriate (less likely to generate spurious sites).
Each ligand is fitted with rigid body refinement to each potential ligand site in the map and the best one for each site selected and written out as a pdb file. The clusters are sorted by size, the biggest one first (with an index of 0). The output placed ligands files have a prefix “best-overall” and are tagged by the cluster index and residue type of the best fit ligand in that site.
By default, the top 10 sites are tested for ligands - to increase this use:
(set-ligand-n-top-ligands 20)
If the “Flexible?” checkbutton is activated, coot will generate a number of variable conformations (default 100) by rotating around the rotatable bonds (torsions). Each of these conformations will be fitted to each of the potential ligand sites in the map and the best one will be selected (again, if it passes the fitting criteria above).
Before you search for flexible ligands you must have read the mmCIF dictionary for that particular ligand residue type (File -> Import CIF dictionary). Use:
(set-ligand-flexible-ligand-n-samples n-samples)
where n-samples is the number of samples of flexibility made for each ligand. Generally speaking, The more the number of rotatable bonds, the bigger this number should be.
By default the options to change these values are not in the GUI. To enable these GUI options, use the scripting function:
(ligand-expert)
After successful ligand searching, one may well want to add that displayed ligand to the current model (the coordinates set that provided the map mask). To do so, use Merge Molecules (Section Merge Molecules).
Sometimes a ligand is placed more or less in the correct position, but the orientation is wrong - or at least you might want to explore other possible orientation. To do that easily a function has been provided:
(flip-ligand imol chain-id residue-number)
This will flip the orientation of the residue around the Eigen vector corresponding to the largest Eigen value, exploring 4 possible orientations.
This function has been further wrapped to provide flipping for the active residue:
(flip-active-ligand)
This function can easily be bound to a key.
As with finding ligands, you are given a chose of maps, protein (masking) atoms. A final selection has to be made for the cut-off level, note that this value is the number of standard deviation of the density of the map before the map has been masked. The default sigma level (water positions must have density above this level) is set for a “2Fo-Fc”-style map. If you want to use a difference map, you must change the sigma level (typically to 3 sigma) otherwise you run the risk of fitting waters to difference map noise peaks.
Then the map is masked by the masking atoms and a search is made of features in the map about the electron density cut-off value. Waters are added if the feature is approximately water-sized and can make sensible hydrogen bonds to the protein atoms. The new waters are optionally created in a new molecule called “Waters”.
You have control over several parameters used in the water finding:
(set-write-peaksearched-waters)
which writes ligand-waters-peaksearch-results.pdb, which
contains the water peaks (from the clusters) without any filtering and
ligand-waters.pdb which are a disk copy filtered waters that
have been either added to the molecule or from which a new molecule
has been created.
(set-ligand-water-to-protein-distance-limits min-d max-d) sets
the minimum and maximum allowable distances between new waters and the
masking molecule (usually the protein). Defaults are 2.4 and 3.2Å.
(set-ligand-water-spherical-variance-limit varlim) sets the
upper limit for the density variance around water atoms. The default
is 0.12.
The map that is marked by the protein and is searched to find the
waters is written out in CCP4 format as "masked-for-waters.map".
Sometimes as a result of water fitting, you may see something like:
WARNING:: refinement failure
start pos: xyz = ( 17.1, 34.76, 60.42)
final pos: xyz = ( 17.19, 34.61, 60.59)
When Coot finds a blob, it does a crude positioning of an atom at the centre of the grid points. It then proceeds to move to the peak of the blob by a series of translations. There are a certain number of cycles, and if it doesn't reach convergence by the end of those cycles then you get the error message.
Often when you go to the position indicated, you can see why Coot had a problem in the refinement.
After a water search, Coot will create a blobs dialog (see Section sec_blobs).
This creates a new residue at the C or N terminal extension of the residue clicked by fitting to the map. \phi,\psi angle pairs are selected at random based on the Ramachandran plot probability (for a generic residue) and fitted to the density. By default there are 100 trials. It is possible that a wrong position will be selected for the terminal residue and if so, you can reject this fit and try again with Fit Terminal Residue 85. Each of the trial positions are scored according to their fit to the map 86 and the best one selected. It is probably a good idea to run “Refine Zone” on these new residues.
If you use the Extensions (Dock Sequence... -> Associate Sequence with Model) to apply a PIR sequence file to a model then Add Terminal Residue will use the sequence alignment to determine the residue type of the added residue.
Sometimes, particularly with low resolution maps, the added terminal residue will wander off to somewhere inappropriate. This can be addressed in a number of ways:
(set-terminal-residue-do-rigid-body-refine 0) will disable
rigid body fitting of the terminal residue fragment for
each trial residue position (the default is 1 (on)) - this may help if
the search does not provide good results.
(set-add-terminal-residue-do-post-refine 1)
(set-add-terminal-residue-n-phi-psi-trials 200) will change the
number of trials (default is 100). This is useful if you think that
Coot needs to search harder to find a good solution to the positioning
of the next residue.
At the C-terminus of a chain of amino-acid residues, there is a “modification” so that the C-O becomes a carbonyl, i.e. an extra (terminal) oxygen (OXT) needs to be added. This atom is added so that it is in the plane of the C\alpha, C and O atoms of the residue.
Scripting usage:
(add-OXT-to-residue imol residue-number insertion-code
chain-id) 87,
where insertion-code is typically "".
Note, in order to place OXT, the N, CA, C and O atoms must be present in the residue - if (for example) the existing carbonyl oxygen atom is called “OE1” then this function will not work.
By default, “Add Atom At Pointer” will pop-up a dialog from which
you can choose the atom type you wish to insert 88. Using
(set-pointer-atom-is-dummy 1) you can by-pass this dialog and
immediately create a dummy atom at the pointer position. Use an
argument of 0 to revert to using the atom type selection
pop-up on a button press.
The atoms are added to a new molecule called “Pointer Atoms”. They should be saved and merged with your coordinates outside of Coot.
The idea is to place a helix more or less “here” (the screen centre) by fitting to the electron density map. The algorithm is straightforward. First we move to the local centre of density, then examine the density for characteristic directions and fit ideal helices (of length 20 residues) to these directions. The helix is then extended if possible (by checking the fit to the map of residues added in ideal helix conformation) and chopped back if not. If the fit is successful, the helix is created in a new molecule called “Helix”. If the fit is not successful, there is instead a message added to the status bar. You can build the majority of a helical protein in a few minutes using this method (you will of course have to assemble the helices and assign residue numbers and sequence later).
This is available as a scripting function (place-helix-here) and
in the GUI (in the “Other Modelling Tools” dialog).
The interface to building ideal polynucleotides can be found by pressing the “Ideal RNA/DNA...” button on the “Other Modelling Tools” dialog.
For a given sequence, a choice of DNA or RNA, A or B form, single or double stranded is presented.
The interface may not gracefully handle uracils in DNA, thymines in RNA or B form RNA 89.
This dialog can be found under “Calculate” in the main menubar. This is typically used to add molecule fragments or residues that are in one molecule to the “working” coordinates 90.
The default temperature factor for new atoms is 30.0. This can be changed by the following
(set-default-temperature-factor-for-new-atoms 50.0)
Let's imagine that you have 3-fold NCS. You have molecule “A” as your master molecule and you make edits to that molecule. Now you want to apply the edits that you made to “A” (the NCS master chain ID) to the “B” and “C” molecules (i.e. you want the “B” and “C” molecules to be rotated/translated versions of the “A” molecule). How is that done?
There are now guis to NCS command to help you out (under Extensions). However, for completeness here are the scripting versions:
(copy-from-ncs-master-to-others imol master-chain-id)
If you have only a range of residues, rather than a whole chain to replace:
(copy-residue-range-from-ncs-master-to-others imol
master-chain-id start-resno end-resno)
e.g.
(copy-residue-range-from-ncs-master-to-others 0 "A" 1 5)
If you want to copy a residue range to a specific chain, or specific list of chains (rather than all NCS peer chains) then make a list of the chain-ids that you wish replaced:
(copy-residue-range-from-ncs-master-to-chains 0 "A" 1 5 (list "C"))
in this case, just the residues in the "C" chain is replaced.
Use the “Run Refmac...” button to select the dataset and the coordinates on which you would like to run Refmac. Note that here Coot only allows the use of datasets which has Refmac parameters set as the MTZ file was read. By default, Coot displays the new coordinates and the new map generated from refmac's output MTZ file. Optionally, you can also display the difference map.
You can add extra parameters
(data lines) to
refmac's input by storing them in a file called
refmac-extra-params in the directory in which you started
coot.
You can also provide extra/replacement parameters for refmac by setting
the variable refmac-extra-params to a list of strings, for
example:
(set! refmac-extra-params (list "REFINE MATRIX 0.1" "MAKE HYDROGENS NO"))
Coot “blocks” 91 until Refmac has terminated 92.
The default refmac executable
is refmac5 it is presumed to be in the
path. If you don't want this, it can be overridden using a
re-definition either at the scripting interface or in one's
~/.coot file e.g.:
(define refmac-exec "/e/refmac-new/bin/refmac5.6.3")
After running refmac several times, you may find that you prefer if the new map that refmac creates (after refmac refinement) is the same colour as the previous one (from before this refmac refinement). If so, use:
(set-keep-map-colour-after-refmac 1)
which will swap the colours of then new and old refmac map so that the post-refmac map has the same colour as the pre-refmac map and the pre-refmac map is coloured with a different colour.
Coot can read shelx .res files and write .ins files, and thus one can refine using SHELXL in a convenient manner using the function
(shelxl-refine imol . hkl-file-name)
(the hkl-file-name is an optional argument)
e.g.
(shelxl-refine 0)
or
(shelxl-refine 0 "insulin.hkl")
In the former case, coot will presume that there is a SHELX hkl file corresponding to the res file that you read in; if there is not coot will print a warning and not try to run shelxl. In the latter case, you can specify the location of the hkl file.
After shelxl has finished, coot will automatically read in the resulting res coordinates, the fcf file, convert the data to mmCIF format and read that, which generates a \sigma_A map and a difference map.
Coot creates a time stamped ins file and a time-stamped sym-link to
the hkl file in the coot-shelxl directory.
Please note that the output ins file will not be particularly useful (and thus shelxl will fail) if the input file was not in SHELX ins format.
There is a GUI for this operation under the “Extensions” menu item.
Sometimes one can click on a button 93 unintentionally. This button is there for such a case. It clears the expectation of an atom pick. This works not only for modelling functions, but also geometry functions (such as Distance and Angle).
Single atoms or residues can be deleted from the molecule using “Delete...” from the “Model/Fit/Refine”dialog. Pressing this button results in a new dialog, with the options of “Residue” (the default), “Atom” and “Hydrogen Atoms”. Now click on an atom in the graphics - the deleted object will be the whole residue of the atom if “Residue” was selected and just that atom if “Atom” was selected. Note that if a residue has an alternative conformation, then “Delete Residue” will delete only the conformation that matches that alternative conformation specifier of the clicked atom.
Only waters are deletable if the "Water" check button is active and waters are not deletable if the "Residue/Monomer" check button is active. This is to reduce mis-clicking.
To rotate the view when in “Delete Mode”, use Ctrl left-mouse.
If you want to delete multiple items you can use check the “Keep Delete Active” check-button on this dialog This will will keep the dialog open, ready for deletion of next item.
You can assign a (FASTA format) sequence to a molecule using:
(assign-fasta-sequence imol chain-id fasta-seq)
This function has been provided as a precursor to functions that will (as automatically as possible) mutate your current coordinates to one that has the desired sequence. It will be used in automatic side-chain assignment (at some stage in the future).
Coot can make an attempt to build missing linking regions or loops 94. This is an area of Coot that needs to be improved, currently O does it much better. We will have several different loop tools here 95. For now there is Calculate -> Fit Gap or the scripting function:
(fit-gap imol chain-id start-resno stop-resno)
and
(fit-gap imol chain-id start-resno stop-resno sequence)
the second form will also mutate and try to rotamer fit the provided sequence.
Example usage: let's say for molecule number 0 in chain "A"
we have residues up to 56 and then a gap after which we have residues
62 and beyond:
(fit-gap 0 "A" 57 61 "TYPWS")
After molecular replacement, the residues of your protein could well have the correct sequence but be chopped back to CG or CB atoms. There is a function to fill such partially-filled residues:
(fill-partial-residues imol)
This identifies residues with missing atoms, then fills them and does a rotamer fit and real-space refinement.
You can change the chain ids of chains using Calculate -> Change Chain IDs.... Coot will block an attempt to change the whole of a chain and the target chain id already exists in the molecule.
If you use the "Residue Range" option then you can insert residues with non-conflicting residue number into pre-existing chains.
As well as the editing “Residue Info” to change occupancies of individual atoms, one can use a scripting function to change occupancies of a whole residue range:
(zero-occupancy-residue-range imol chain-id
resno-start resno-last)
example usage:
(zero-occupancy-residue-range 0 "A" 23 28)
This is often useful to zero out a questionable loop before submitting for refinement. After refinement (with refmac) there should be relatively unbiased density in the resulting 2Fo-Fc-style and difference maps.
Similarly there is a function to reverse this operation:
(fill-occupancy-residue-range imol chain-id
resno-start resno-last)
Currently this is available only in scripting form:
(fix-nomenclature-errors imol)
This will fix atoms nomenclature problems in molecule number imol according to the same criteria as WATCHECK 96 e.g. Chi-2 for Phe, Tyr, Asp, and Glu should be between -90 and 90 degrees. Note that Val and Leu nomenclature errors are also corrected.
There is an experimental scripting function
(fit-protein imol)
which does a auto-fit rotamer and Real Space Refinement for each residue. The graphics follow the refinement.
All the waters in a model can be refined (that is, moved to the local density peak) using
(fit-waters imol)
This is a non-interactive function (the waters are moved without user intervention).
Often you want to move a ligand (or some such) from wherever it was read in to the position of interest in your molecule (i.e. the current view centre). There is a GUI to do this: Calculate -> Move Molecule Here.
There are scripting functions available for this sort of thing:
(molecule-centre imol)
will tell you the molecule centre of the imolth molecule.
(translate-molecule-by imol x-shift y-shift z-shift)
will translate all the atoms in molecule imol by the
given amount (in Ångströms).
(move-molecule-to-screen-centre imol)
will move the imolth molecule to the current centre of the screen (sometimes useful for imported ligands). Note that this moves the atoms of the molecule - not just the view of the molecule.
If you don't like the labels "Rotate/Translate Zone" or "Place Atom at Pointer" and rather they said something else, you can change the button names using:
(set-model-fit-refine-rotate-translate-zone-label "Move Zone")
and
(set-model-fit-refine-place-atom-at-pointer "Add Atom")
Maps are “infinite,” not limited to pre-calculated volume (the “Everywhere You Click - There Is Electron Density” (EYC-TIED) paradigm) symmetry-related electron density is generated automatically. Maps are easily re-contoured. Simply use the scroll wheel on you mouse to alter the contour level (or -/+ on the keyboard). Maps follow the molecule. As you recentre or move about the crystal, the map quickly follows. If your computer is not up to re-contouring all the maps for every frame, then use Draw -> Dragged Map... to turn off this feature.
Unfortunately, there is a bug in map-reading. If the map is not a bona-fide CCP4 map 97, then coot will crash. Sorry. A fix is in the works but “it's complicated”. That's why maps are limited to the extension ".ext" and ".map", to make it less likely a non-CCP4 map is read.
From MTZ, mmCIF and .phs data use File -> Open MTZ, CIF or phs.... You can then choose the MTZ columns for the Fourier synthesis. The button “Expert mode” also adds to the options any anomalous columns you may have in the MTZ file (a -90 degree phase shift will be applied). It also provides the option to apply resolution limits.
From a CCP4 map use File -> Read Map. After being generated/read, the map is immediately contoured and centred on the current rotation centre.
This function allows Coot to read an MTZ file and make a map directly (without going through the column selection procedure). The default column labels for auto-reading are "FWT" and "PHWT" for the 2Fo-Fc-style map, "DELFWT" and "PHDELWT" for the difference map. You can change the column labels that Coot uses for auto-reading - here is an example of how to do that:
(set-auto-read-column-labels "2FOFCWT" "PHIWT" 0)
(set-auto-read-column-labels "FOFCWT" "DELPHIWT" 1)
By default the difference map is created in auto-reading the MTZ file. If you don't want a difference map, you can use the function:
(set-auto-read-do-difference-map-too 0)
There are several maps that can be generated from CIF files that contain observed Fs, calculated Fs and calculated phases:
(read-cif-data-with-phases-fo-alpha-calc
cif-file-name) Calculate an atom map using F_obs and
\alpha_calc
(read-cif-data-with-phases-2fo-fc cif-file-name)
Calculate an atom map using F_obs, F_calc and
\alpha_calc
(read-cif-data-with-phases-fo-fc cif-file-name)
Calculate an difference map using F_obs, F_calc and
\alpha_calc.
There are 2 ways to read data by scripting:
(read-phs-and-make-map-using-cell-symm phs-file-name space-group-name a b c alpha beta gamma)
(read-pdb-and-make-map-with-reso-limits imol-previous phs-file-name reso-limit-low reso-limit-high)
The first specifies the cell explicitly, and alpha, beta and gamma are specified in degrees.
The second form allows the specification of resolution limits and takes the cell and symmetry from a previous molecule (typically a pdb file).
Maps can be re-contoured using the middle-mouse scroll-wheel (buttons 4 and 5 in X Window System(TM) terminology). Scrolling the mouse wheel will change the map contour level and the map it redrawn. If you have several maps displayed then the map that has its contour level changed can be set using HID -> Scrollwheel -> Attach scroll-wheel to which map?. If there is only one map displayed, then that is the map that has its contour level changed (no matter what the scroll-wheel is attached to in the menu). The level of the electron density is displayed in the top right hand corner of the OpenGL canvas.
Use keyboard <+> or <-> to change the contour level if you don't have a scroll-wheel 98.
If you are creating your map from an MTZ file, you can choose to click on the “is difference map” button on the Column Label selection widget (after a data set filename has been selected) then this map will be displayed in 2 colours corresponding to + and - the map contour level.
If you read in a map and it is a difference map then there is a checkbutton to tell Coot that.
If you want to tell Coot that a map is a difference map after it has been read, use:
(set-map-is-difference-map imol)
where imol is the molecule number.
By default the change of the contour level is determined from the sigma of the map. You can change this in the map properties dialog or by using the scripting function:
(set-contour-by-sigma-step-by-mol step on/off? imol)
where
step is the difference in sigma from one level to the next (typically 0.2)
on/off? is either 0 (sigma stepping off) or 1 (sigma stepping on)
By default the map radius 99 is 10Å. The default increment to the electron density depends on whether or not this is a difference map (0.05 e^-/\AA^3 for a “2Fo-Fc” style map and 0.005 e^-/\AA^3 for a difference map). You can change these using Edit -> Map Parameters or by using the “Properties” button of a particular map in the Display Control (Display Manager) window.
The extent of the map can be set using the GUI (Edit -> Map Parameters -> Map Radius) or by using the scripting function, e.g.:
(set-map-radius 13.2)
Usually one doesn't want to look at negative contour levels of a map100, so Coot has by default a limit that stops the contour level going beyond (less than) 0. To remove the limit:
(set-stop-scroll-iso-map 0) for a 2Fo-Fc style map
(set-stop-scroll-diff-map 0) for a difference map
To set the limits to negative (e.g. -0.6) levels:
(set-stop-scroll-iso-map-level -0.6)
and similarly:
(set-stop-scroll-diff-map-level -0.6)
where the level is specified in e^-/\AA^3.
The width of the lines that describe the density can be changed like this:
(set-map-line-width 2)
The default line width is 1.
By default, maps get coloured according to their molecule number. The starting colour (i.e. for molecule 0) is blue. The colour of a map can be changed by Edit -> Map Colour... The map colour gets updated as you change the value in the colour selector 101. Use “OK” to fix that colour.
As subsequent maps are read, they are coloured by rotation round a colour wheel. The default colour map step is 31 degrees. You can change this using:
(set-colour-map-rotation-for-map step)
For some strange reason, some crystallographers 102 like to have their difference maps coloured with red as positive and green as negative, this option is for them:
(set-swap-difference-map-colours 1)
This option will allow the “blue is positive, red is negative” colour scheme on “Edit -> Map Colour”.
Using the “Make a Difference Map” function in the Extensions menu, one can make a difference from two arbitrary maps. The maps need not be on the same griding, or in the same space group even. The resulting map will be on the same griding and space group as the “Reference” map.
There is a scripting interface to the generation of map averages. As above, the maps need not be on the same grid or in the same space group. The resulting map will have the same gridding and space group as the first map in the list. Typical usage:
(average-map '((1 1.0) (2 1.0))))
The argument to (average-map is a list of lists, each list
element is a list of the map number and a weighting factor (1.0 in this
case).
By default, the Shannon sampling factor is the conventional 1.5. Use larger values (Edit -> Map Parameters -> Sampling Rate) for smoother maps 103.
This value can be set by the scripting command
(set-map-sampling-rate 2.5)
By default, the map is re-contoured at every frame during a drag (Ctrl Left-mouse). Sometimes this can be annoyingly slow and jerky so it is possible to turn it off: Draw -> Dragged Map -> No.
To change this by scripting:
(set-active-map-drag-flag 0)
If activated (Edit -> Map Parameters -> Dynamic Map Sampling) the map will be re-sampled on a more coarse grid when the view is zoomed out. If “Display Size” is also activated, the box of electron density will be increased in size also. In this way, you can see electron density for big maps (many unit cells) and the graphics still remain rotatable.
If you want to have these functions active for all maps, add the following to your initialization file Scheme:
(set-dynamic-map-sampling-on)
(set-dynamic-map-size-display-on)
The skeleton (also known as “Bones” 104) can be displayed for any map. A map can be skeletonized using Calculate -> Map Skeleton.... Use the option menu to choose the map and click “On” then “OK” to the generate the map (the skeleton is off by default).
The level of the skeleton can be changed by using Edit -> Skeleton Parameters... -> Skeletonization Level... and corresponds to the electron density level in the map. By default this value is 1.2 map standard deviations. The amount of map can be changed using Edit -> Skeleton Parameters... -> Skeleton Box Radius...105. The units are in Ångströms, with 40 as the default value.
The skeleton is often recalculated as the screen centre changes - but not always since it can be an irritatingly slow calculation. If you want to force a regeneration of the displayed skeleton, simply centre on an atom (using the middle mouse button) or press the <S> key.
It can be educational (even useful at lower resolutions) to sharpen or blur a map. This can be achieved with the sharpening tool Calculate -> Map Sharpening.... By default, the maximum and minimum sharpness is +/- 30Å^2, this can be changed (in this case to 80) using:
(set-map-sharpening-scale-limit 80)
This currently only works on maps created by reading an MTZ (or other) reflection data file.
Pattersons can be generated using the make-and-draw-patterson
function. Example usage:
(make-and-draw-patterson mtz-file-name f-col sig-f-col weight-col use-weights-flag)
where use-weights-flag is either 0 or 1.
A map can be masked by a set of coordinates. Use the scripting function:
(mask-map-by-molecule imol-map imol-model invert-mask?)
If invert-mask? is 0, this will create a new map that has density only where there are no (close) coordinates. If invert-mask? is 1 then the map density values will be set to zero everywhere except close to the atoms of molecule number imol-model.
The radius of the mask around each atom is 2.0Å by default. You can change this using:
(set-map-mask-atom-radius radius)
There is a GUI interface to Map Masking under the Extensions menu.
If one wanted to show just the density around a ligand:
(mask-map-by-molecule 2 1 1)
This creates a new map. Turn the other maps off, leaving only the masked map.
To get a nice rendered image, press F8 (see Section Raster3D).
If you want to remove all the atoms 106 that lie “outside the map” (i.e. in low density) you can use
(trim-molecule-by-map imol-coords imol-map density-level delete/zero-occ?)
where delete/zero-occ? is 0 to remove the atoms and
1 to set their occupancy to zero.
There is a GUI interface for this feature under the “Extensions” menu item.
If you want to transform a map, you can do it thusly:
(transform-map imol rotation-matrix trans point radius)
where:
rotation-matrix is a 9-membered list of numbers for an orthogonal rotation matrix.
trans is a 3-membered list of numbers (distances in Ångstöms).
point is a 3-membered list of numbers (centre point in Ångstöms).
radius is a single number (also in Ångstöms).
This applies the rotation rotation-matrix and a translation trans to a map fragment, so that when the transformation is applied the centre of the new map is at point.
Example usage:
(transform-map 2 '(1 0 0 0 1 0 0 0 1) '(0 0 1) (rotation-centre) 10)
which transforms map number 2 by a translation of 1Å along the Z axis, centred at the screen centre for 10Å around that centre.
Here's a more real-world example:
Let's say we want to tranform the density over the “B” molecule to a position over the “A” molecule. First we do a LSQ transformation to get the rotation and translation that moves the “B” coordinates over the “A” coordinates:
In the terminal output we get:
| 0.9707, 0.2351, 0.05033| | -0.04676, 0.39, -0.9196| | -0.2358, 0.8903, 0.3896| ( -33.34, 21.14, 18.82)
The centre of the “A” molecule is at (58.456, 5.65, 11.108). So we do:
(transform-map 3 (list 0.9707 0.2351 0.05033 -0.04676 0.39 -0.9196
-0.2358 0.8903 0.3896) (list -33.34 21.14 18.82) (list 58.456 5.65
11.108) 8)
Which creates a map over the middle of the “A” molecule. Note that using a too high radius can cause overlap problems, so try with a small radius (e.g. 5.0) if the resulting map looks problematic.
Alternatively, instead of typing the whole matrix, you can use a
coordinates least-squares fit to generate the matrix for you.
(transform-map-using-lsq-matrix) does just that.
Heres how to use it:
(transform-map-using-lsq-matrix imol-ref ref-chain
ref-resno-start ref-resno-end imol-mov mov-chain mov-resno-start
mov-resno-end imol-map about-pt radius)
Hopefully the arguments are self explanatory (ref refers
to the reference molecule, of course and about-pt is a
3-number list such as is returned by (rotation-centre)).
We can now export that map, if we want.
You can write out a map from Coot (e.g. one from NCS averaging, or masking or general transformation) using the export map function:
(export-map imol filename)
e.g.
(export-map 4 "ncs-averaged.map")
The validation functions are still being added to from time to time. In future there will be more functions, particularly those that will interface to other programs.
Ramachandran plots are “dynamic”. When you edit the molecule (i.e. move the coordinates of some of atoms) the Ramachandran plot gets updated to reflect those changes. Also the underlying \phi/\psi probability density changes according to the selected residue type (i.e. the residue under the mouse in the plot). There are 3 different residue types: GLY, PRO, and not-GLY-or-PRO 107.
When you mouse over a representation of a residue (a little square or triangle 108) the residue label pops up. The residue is “active” i.e. it can be clicked. The “graphics” view changes so that the C\alpha of the selected residue is centred. In the Ramachandran plot window, the current residue is highlighted by a green square.
The underlying distributions are taken from the Richardson's Top500 structures http://kinemage.biochem.duke.edu/databases/top500.php.
The probability levels for acceptable (yellow) and preferred (red) are 0.2% and 2% respectively.
You can change the contour levels:
(set-ramachandran-plot-contour-levels 0.025 0.003)
You can change the “blocksize” (the default is 10 degrees) of the contours using
(set-ramachandran-plot-background-block-size 5)
These comes into effect when a new plot is created (it doesn't change plots currently displayed).
A restraints-based geometry analysis of the molecule. The distortion is weighted by atom occupancy. The distortion of the geometry due to links is shared between the contributing residues.
Note that only the first model of a multi-model molecule is analysed.
The dictionary is used to identify the chiral atoms of each of the model's residues. A clickable list is created of atoms whose chiral volume in the model is of a different sign to that in the dictionary.
During refinement and regularization, Coot will pop-up dialogs warning about chiral volume errors - if you have them. This can be annoying 109. You can inhibit this dialog like this:
(set-show-chiral-volume-errors-dialog 0)
There are two obvious ways:
1) mutate and auto-fit rotamer (mutate it to the residue type that it is)
2) RS Refine the residue and invert the chiral centre by pulling an atom. Usually you can pull the CA to the other side of the plane made by the chiral neighbouring atoms (using ctrl left-click). Sometimes giving the CB a good old tweak is the easier way.
Inverting the CB of THR is easier, just move the OG so that the plane of the neighbours is on the other side of the CB (again with ctrl left-click).
This is an interface to the Blobs dialog. A map and a set of coordinates that model the protein are required.
A blob is region of relatively high residual election density that cannot be explained by a simple water. So, for example, sulfates, ligands, mis-placed sidechains or unbuilt terminal residues might appear as blobs. The blobs are in order, the biggest 110 at the top.
This is one of the fastest ways to validate a model and its data (presuming that the difference map comes from a post-refinement mFo-DFc map). It highlights regions where the model and the data do not agree.
Lesser peaks within a certain distance (by default, 2.0Å) of a large peak are not shown. This cuts down on the number of times one is navigated to a particular region because of ripple or other noise peaks around a central peak.
This value can be queried:
(difference-map-peaks-max-closeness)
and adjusted:
(set-difference-map-peaks-max-closeness 0.1)
Sometimes waters can be misplaced - taking the place of sidechains or ligands or crystallization agents such as phosphate for example 111. In such cases the variance of the difference map can be used to identify these problems.
This function is also useful to check anomalous maps. Often waters are placed in density that is really a something else, perhaps a cation, anion, sulphate or a ligand. If such an atom diffracts anomalously this can be identified and corrected.
By default the waters with a map variance greater than 3.5\sigma are listed. One can be more rigorous by using a lower cut-off:
(set-check-waters-by-difference-map-sigma-level 3.0)
The scripting interface is:
(check-waters-by-difference-map imol-coords
imol-diff-map)
where imol-coords is the molecule number of the coordinates that contain the waters to be checked
imol-diff-map is the molecule number of the difference map (it must be a difference map, not an “ordinary” map). This difference map must have been calculated using the waters. So there is no point in doing this check immediately after “Find Waters”. You will need to run Refmac or some other refinement first first 112.
The molprobity tools probe and reduce have been interfaced into Coot (currently, the interface is not as slick as it might be). However, the tools are useful and can be used in the following way:
first we need to tell Coot where to find the relevant executables (typically you would add the following lines to you ~/.coot file):
(define *probe-command* "/path/to/probe/executable")
(define *reduce-command* "/path/to/reduce/executable")
now the probe hydrogens and probe dots can be generated using Validate -> Probe Clashes (or in the Scripting Window):
(probe imol)
where imol is the molecule number of coordinates to be probed. A new molecule with Hydrogens is created (by reduce) and read in.
By default Coot creates a new molecule for the molecule that now has hydrogens. To change this:
(set! reduce-molecule-updates-current #t)
and that, as you can guess, replaces, rather than adds to the “probed” molecule.
This gives a "static" view of the molecule's interactions.
To get a dynamic view (which is currently only enabled on rotating chi angles) add these to your ~/.coot file:
(set-do-probe-dots-on-rotamers-and-chis 1)
To get a semi-static view (dots are regenerated in the region of zone after a "Real Space Refinement"):
(set-do-probe-dots-post-refine 1)
It is often difficult to detect by eye the correct orientation of the amino-carbonylo group of GLN and ASNs. However, we can use (properly refined) temperature factors to detect outliers. We take the Z value as half the difference between the B-factor of the NE2 and OE1 divided by the standard deviation of the B-factors of the rest of the residue. An analysis of GLNs and ASNs of high resolutions structures indicates that a Z value of greater than 2.25 indicates a potential (if not probable) flip. A “Fix” button is provided in the resultant dialog make this easy to do.
This analysis was added after discussions with Atsushi Nakagawa and so is called “Nakagawa's Bees”.
The analysis does not check residues with multiple conformations.
Coot provides several graphs that are useful for model validation (on a residue by residue basis): residue density fit, geometry distortion, temperature factor variance, peptide distortion and rotamer analysis.
The density fit graph shows the density fit for residues. The score is the average electron density level at the atom centres of the atoms in the residue. The height of the blocks is inversely proportional to the density average.
The residue density fit is by default scaled to a map that is calculated on the absolute scale. Sometimes you might be using a map with density levels considerably different to this, which makes the residue density fit graph less useful. To correct for this you can use the scripting function:
(set-residue-density-fit-scale-factor factor)
where factor would be 1/(4\sigma_map) (as a rule of thumb).
(residue-density-fit-scale-factor) returns the current scale
factor (default 1.0).
There is also a GUI to this:
Extensions -> Refine... -> Set Density Fit Graph Weight...
Residue rotamers are scored according to the prior likelihood. Note that when CD1 and CD2 of a PHE residue are exchanged (simply a nomenclature error) this can lead to large red blocks in the graph (apparently due to very unlikely rotamers). There are several other residues that can have nomenclature errors like this. To fix these problems use
(fix-nomenclature-errors imol)
This idea is from Eleanor Dodson, who liked to use the standard deviation of a residue's temperature factors to highlight regions of questionable structure.
Note that Hydrogens are ignored in this analysis.
Some variability of the \omega is to be expected in the peptide bond. But not too much. Anything more than 13 degrees is suspicicous. Unexpected peptide bonds show up red by default. If cis peptides are to be expected, and should not marked as bad, then you can tell this to Coot using:
Edit -> Preferences -> Geometry -> Cis-Peptides -> No
Coot uses the surface code from Gruber and Noble (2004).
Coot uses the partial charges of the atoms (the partial_charge field in the _chem_comp_atom block) from the charge dictionary item in the refmac (or other) cif dictionary. However, partial charges are only used under certain conditions
1) the molecule consists of less than 100 atoms
or
2) the number of atoms in the molecule that are hydrogens is at least 15% of the total number of atoms in the molecule
If partial charges are not used, then the fall-back is to use charges from side-chains charged at physiological pH (Arg, Lys, Asp, Glu).
This manual is on the web where it can be searched:
In the Menu item “About”, under “Online Docs URL...” there is a entry bar that can be used to search the Coot documentation via Google. The results are returned as a web page in web browser. The browser type can be specified as in this example:
(set-browser-interface "firefox")
Example usage can be found in xxx/share/coot/scheme/group-settings.scm
Building structures using low resolution data is a pain. We hope to make it less of a pain in future, but there are some things that you can do now.
(set-matrix 20.0)
[Default is 60, the lower the number the more the geometry is idealised]
This describes the files and directory that coot leaves behind after it has been fed (sorry, I mean “used”). Everything except the 0-coot.state.scm state file can comfortably be deleted if needed after coot has finished.
You can stop the state and history files being written if you start coot
with the --no-guano option.
Coot will occasionally ask you to clear up the coot-backup directory. You can adjust the behaviour in a number of ways:
(define *clear-out-backup-run-n-days* 3) will run the backup clearance every 3 days (the default is every 7).
(define *clear-out-backup-old-days* 1) will clear out files older then 1 day (rather than the default 7 days).
(clear-backups-maybe)
So, if you wanted to clear out everything more than 1 day old, every time, without Coot asking you about it:
(define *clear-out-backup-run-n-days* 0)
(define *clear-out-backup-old-days* 1)
(define (clear-backups-maybe)
(delete-coot-backup-files 'delete)
(coot-real-exit 0))
If you get stuck in "translate" mode in the GL canvas (i.e. mouse does not rotate the view as you would expect) simply press and release the Ctrl key to return to "rotate" mode.
The keyboard <I> key toggles the “continuous rotation” mode. The menu item Draw -> Spin View On/Off does the same thing.
Similarly, if you are stuck in a mode where the “Model/Fit/Refine” buttons don't work (the atoms are not selected, only the atom gets labelled), press and release the Shift key.
Button labels ending in “...” mean that a new dialog will pop-up when this button is pressed.
Note that left-mouse in the graphics window is used for both atom picking and rotating the view, so try not to click over an atom when trying to rotate the view when in atom selection mode.
Click and drag using right-mouse (up and down or left and right) to zoom in and out.
To change the map to which the scroll-wheel is attached, use the scroll check button in the Display Manager or use HID -> Scrollwheel -> Attach Scrollwheel to which map?
Several of the parameters of Coot are chosen because they are reasonable on my “middle-ground” development machine. However, these parameters can be tweaked so that slower computers perform better:
(set-smooth-scroll-steps 4) ; default 8
(set-smooth-scroll-limit 30) ; Angstroms
(set-residue-selection-flash-frames-number 3);
(set-skeleton-box-size 20.0) ; A (default 40).
(set-active-map-drag-flag 0) ; turn off recontouring every step
(set-idle-function-rotate-angle 1.5) ; continuous spin speed
findligand is a stand-alone command-line program that uses the libraries of Coot.
findligand provides a number of command line arguments for increased flexibility:
--pdbin pdb-in-filename
where pdb-in-filename is the protein (typically)
--hklin mtz-filename
--f f_col_label
--phi phi_col_label
--clusters nclust
where nclust is the number of density clusters (potential ligand sites) to search for
--sigma sigma-level
where sigma-level the density level (in sigma) above which the map is searched for ligands
--fit-fraction frac
where frac is the minimum fraction of atoms in density allowed after fit [default 0.75]
--flexible
means use torsional conformation ligand search
--samples nsamples
nsamples is the number of flexible conformation samples [default 30]
--dictionary cif-dictionary-name
the file containing the CIF ligand dictionary description
One uses findligand like this:
$ findligand various-args ligand-pdb-file-name(s)
|
i.e. the example ligand pdb files that you wish to search for are given at the end of the command line.
Where mode is an integer number
How should the mouse move the view?
mode=1 for "Flat", mode=2 for "Spherical Surface"
return the mouse view status mode
mode=1 for "Flat", mode=2 for "Spherical Surface"
tell coot that you prefer to run python scripts if/when there is an option to do so.
Where dir is a string
make a directory dir (if it doesn't exist) and return error code
If it can be created, create the directory dir, return the success status like mkdir: mkdir
Returns: zero on success, or -1 if an error occurred. If dir already exists as a directory, return 0 of course.
Where i is an integer number
Show Paths in Display Manager?
Some people don't like to see the full path names in the display manager here is the way to turn them off, with an argument of 1.
return the internal state
What is the internal flag?
Returns: 1 for "yes, display paths" , 0 for not
Where ext is a string
add an extension to be treated as coordinate files
Where ext is a string
add an extension to be treated as data (reflection) files
Where ext is a string
add an extension to be treated as geometry dictionary files
Where ext is a string
add an extension to be treated as geometry map files
Where ext is a string
remove an extension to be treated as coordinate files
Where ext is a string
remove an extension to be treated as data (reflection) files
Where ext is a string
remove an extension to be treated as geometry dictionary files
Where ext is a string
remove an extension to be treated as geometry map files
sort files in the file selection by date?
some people like to have their files sorted by date by default
do not sort files in the file selection by date?
removes the sorting of files by date
Where istate is an integer number
on opening a file selection dialog, pre-filter the files.
set to 1 to pre-filter, [0 (off, non-pre-filtering) is the default
Where txt is a string
create a dialog with information
create a dialog with information string txt. User has to click to dismiss it, but it is not modal (nothing in coot is modal).
Where filename is a string
given a filename, try to read it as a data file
We try as .phs and .cif files first
Where:
- chain_id is a string
- imol is an integer number
the number of residues in chain chain_id and molecule number imol
Returns: the number of residues
Where:
- imol is an integer number
- chain_id is a string
- serial_num is an integer number
return the rename from a residue serial number
Returns: NULL (scheme False) on failure.
Where:
- imol is an integer number
- chain_id is a string
- serial_num is an integer number
a residue seqnum (normal residue number) from a residue serial number
Returns: < -9999 on failure
Where:
- imol is an integer number
- chain_id is a string
- serial_num is an integer number
the insertion code of the residue.
Returns: NULL (scheme False) on failure.
Where imol is an integer number
the chain_id (string) of the ichain-th chain molecule number imol
return the number of models in molecule number imoluseful for NMR or other such multi-model molecules.
return the number of models or -1 if there was a problem with the given molecule.
Returns: the chain-id
Where imol is an integer number
number of chains in molecule number imol
Returns: the number of chains
Where:
- imol is an integer number
- chain_id is a string
is this a solvent chain? [Raw function]
This is a raw interface function, you should generally not use this, but instead use (is-solvent-chain? imol chain-id)
Returns: -1 on error, 0 for no, 1 for is "a solvent chain". We wouldn't want to be doing rotamer searches and the like on such a chain.
Where imol is an integer number
return the number of residues in the molecule,
return -1 if this is a map or closed.
Where imol is an integer number
sort the chain ids of the imol-th molecule in lexographical order
Where imol is an integer number
a gui dialog showing remarks header info (for a model molecule).
Where imol is an integer number
simply print secondardy structure info to the terminal/console. In future, this could/should return the info.
Where imol is an integer number
copy molecule imol
Returns: the new molecule number. Return -1 on failure to copy molecule (out of range, or molecule is closed)
Where:
- imol_ligand_new is an integer number
- chain_id_ligand_new is a string
- resno_ligand_new is an integer number
- imol_current is an integer number
- chain_id_ligand_current is a string
- resno_ligand_current is an integer number
Copy a molecule with addition of a ligand and a deletion of current ligand.
This function is used when adding a new (modified) ligand to a structure. It creates a new molecule that is a copy of the current molecule except that the new ligand is added and the current ligand/residue is deleted.
Where imol is an integer number
Experimental interface for Ribosome People.
Ribosome People have many chains in their pdb file, they prefer segids to chainids (chainids are only 1 character). But coot uses the concept of chain ids and not seg-ids. mmdb allow us to use more than one char in the chainid, so after we read in a pdb, let's replace the chain ids with the segids. Will that help?
the coot version string
Returns: something like "coot-0.1.3". New versions of coot will always be lexographically greater than previous versions.
return the subversion revision number of this build.
Used in finding updates.
Where imol is an integer number
return the name of molecule number imol
Returns: 0 if not a valid name ( -> False in scheme) e.g. "/a/b/c.pdb" for "d/e/f.mtz FWT PHWT"
Where:
- imol is an integer number
- new_name is a string
set the molecule name of the imol-th molecule
Where retval is an integer number
exit from coot, give return value retval back to invoking process.
What is the molecule number of first coordinates molecule?
return -1 when there is none.
molecule number of first small (<400 atoms) molecule.
return -1 on no such molecule
What is the molecule number of first unsaved coordinates molecule?
return -1 when there is none.
Where state is an integer number
set the bond lines to be antialiased
Where state is an integer number
turn the GL lighting on (state = 1) or off (state = 0)
slows down the display of simple lines
shall we start up the Gtk and the graphics window?
if passed the command line argument –no-graphics, coot will not start up gtk itself.
An interface function for Ralf.
start Gtk (and graphics)
This function is useful if it was not started already (which can be achieved by using the command line argument –no-graphics).
An interface for Ralf
"Reset" the view
return 1 if we moved, else return 0.
centre on last-read molecule with zoom 100. If we are there, then go to the previous molecule, if we are there, then go to the origin.
return the number of molecules (coordinates molecules and map molecules combined) that are currently in coot
Returns: the number of molecules (closed molecules are not counted)
Where:
- width_scale is a number
- frequency_scale is a number
Settings for the inevitable discontents who dislike the default rocking rates (defaults 1 and 1).
Where f is a number
how far should we rotate when (auto) spinning? Fast computer? set this to 0.1
Where filename is a string
a synonym for read-pdb. Read the coordinates from filename (can be pdb, cif or shelx format)
Where state is an integer number
shall we convert nucleotides to match the dictionary names?
Usually we do not want to do this (give current Coot architecture). Most often not, though. Coot should handle the residue synonyms transparently.
default off (0).
Where:
- filename is a string
- recentre_on_read_pdb_flag is an integer number
read coordinates from filename with option to not recentre.
set recentre_on_read_pdb_flag to 0 if you don't want the view to recentre on the new coordinates.
Where filename is a string
read coordinates from filename and recentre the new molecule at the screen rotation centre.
Where imol is an integer number
some programs produce PDB files with ATOMs where there should be HETATMs. This is a function to assign HETATMs as per the PDB definition.
Where:
- imol is an integer number
- chain_id is a string
- resno is an integer number
- ins_code is a string
if this is not a standard group, then turn the atoms to HETATMs.
Return 1 on atoms changes, 0 on not. Return -1 if residue not found.
Where:
- imol is an integer number
- chain_id is a string
- resno is an integer number
- ins_code is a string
residue has HETATMs?
return 1 if all atoms of the specified residue are HETATMs, else, return 0. If residue not found, return -1.
Where:
- imol_target is an integer number
- imol_fragment is an integer number
- atom_selection is a string
replace the parts of molecule number imol that are duplicated in molecule number imol_frag
Where:
- imol_target is an integer number
- chain_id_target is a string
- imol_reference is an integer number
- chain_id_reference is a string
- resno_range_start is an integer number
- resno_range_end is an integer number
copy the given residue range from the reference chain to the target chain
resno_range_start and resno_range_end are inclusive.
Where:
- molecule_number is an integer number
- file_name is a string
replace pdb. Fail if molecule_number is not a valid model molecule. Return -1 on failure. Else return molecule_number
Where filename is a string
dump the current screen image to a file. Format ppm
You can use this, in conjunction with spinning and view moving functions to make movies
Where:
- imol is an integer number
- state is an integer number
sets the density map of the given molecule to be drawn as a (transparent) solid surface.
Where:
- imol is an integer number
- state is an integer number
toggle for standard lines representation of map.
This turns off/on standard lines representation of map. transparent surface is another representation type.
If you want to just turn off a map, don't use this, use
.
Where:
- imol is an integer number
- opacity is a number
set the opacity of density surface representation of the given map.
0.0 is totally transparent, 1.0 is completely opaque and (because the objects are no longer depth sorted) considerably faster to render. 0.3 is a reasonable number.
Where state is an integer number
set the flag to do flat shading rather than smooth shading for solid density surface.
Default is 1 (on.
Where istate is an integer number
Some people (like Phil Evans) don't want to scroll their map with the mouse-wheel.
To turn off mouse wheel recontouring call this with istate value of 0
return the internal state of the scroll-wheel map contouring
Where n_sigma is a number
set the default inital contour for 2FoFc-style map
in sigma
Where n_sigma is a number
set the default inital contour for FoFc-style map
in sigma
Where element is an integer number
internal function to get an element of the view quaternion. The whole quaternion is returned by the scheme function view-quaternion
Where:
- i is a number
- j is a number
- k is a number
- l is a number
Set the view quaternion.
Where:
- imol is an integer number
- current_chain is a string
- next_ncs_chain is a string
Given that we are in chain current_chain, apply the NCS operator that maps current_chain on to next_ncs_chain, so that the relative view is preserved. For NCS skipping.
Where:
- imol is an integer number
- current_chain is a string
- next_ncs_chain is a string
- forward_flag is an integer number
as above, but shift the screen centre also.
Where istate is an integer number
set a flag: is the origin marker to be shown? 1 for yes, 0 for no.
show the vertical modelling toolbar in the GTK2 version (the toolbar is shown by default)
show all available icons in the modelling toolbar (same as MFR dialog)
Where state is an integer number
to swap sides of the Model/Fit/Refine toolbar 0 (default) is right, 1 is left, 2 is top, 3 is bottom
reparent the Model/Fit/Refine dialog so that it becomes part of the main window, next to the GL graphics context
Where s is a string
Put text s into the status bar.
use this to put info for the user in the statusbar (less intrusive than popup).
Where state is an integer number
Alternate mode for rotation.
Prefered by some, including Dirk Kostrewa. I don't think this mode works properly yet
Put the blob under the cursor to the screen centre. Check only positive blobs. Useful function if bound to a key.
The refinement map must be set. (We can't check all maps because they are not (or may not be) on the same scale).
Returns: 1 if successfully found a blob and moved there. return 0 if no move.
Where icursor_index is an integer number
let the user have a different pick cursor
sometimes (the default) GDK_CROSSHAIR is hard to see, let the user set their own
Where txt is a string
Allow the changing of Model/Fit/Refine button label from "Rotate/Translate Zone".
Where txt is a string
Allow the changing of Model/Fit/Refine button label from "Place Atom at Pointer".
Where state is an integer number
shall atoms with zero occupancy be moved when refining? (default 1, yes)
return the state of "shall atoms with zero occupancy be moved when refining?"
Where imol is an integer number
turn off backups for molecule number imol
Where imol is an integer number
turn on backups for molecule number imol
Where imol is an integer number
return the backup state for molecule number imol
return 0 for backups off, 1 for backups on, -1 for unknown
Where imol is an integer number
set the molecule number imol to be marked as having unsaved changes
Where imol is an integer number
does molecule number imol have unsaved changes?
Returns: -1 on bad imol, 0 on no unsaved changes, 1 on has unsaved changes
Where imol is an integer number
set the molecule to which undo operations are done to molecule number imol
show the Undo Molecule chooser - i.e. choose the molecule to which the "Undo" button applies.
Where state is an integer number
set the state for adding paths to backup file names
by default directories names are added into the filename for backup (with / to _ mapping). call this with state=1 to turn off directory names
Where state is an integer number
set if backup files will be compressed or not using gzip
recover session
After a crash, we provide this convenient interface to restore the session. It runs through all the molecules with models and looks at the coot backup directory looking for related backup files that are more recent that the read file. (Not very good, because you need to remember which files you read in before the crash - should be improved.)
Where mtz_file_name is a string
fire up a GUI, which asks us which model molecule we want to calc phases from. On "OK" button there, we call map_from_mtz_by_refmac_calc_phases()
Where:
- mtz_file_name is a string
- f_col is a string
- sigf_col is a string
- imol_coords is an integer number
Calculate SFs (using refmac optionally) from an MTZ file and generate a map. Get F and SIGF automatically (first of their type) from the mtz file.
Returns: the new molecule number, -1 on a problem.
Where:
- mtz_file_name is a string
- f_col is a string
- sigf_col is a string
- imol_coords is an integer number
Calculate SFs from an MTZ file and generate a map.
Returns: the new molecule number.
Where imap is an integer number
set the map that is moved by changing the scroll wheel and change_contour_level().
Where imol is an integer number
return the molecule number to which the mouse scroll wheel is attached
set the map that has its contour level changed by the scrolling the mouse wheel to molecule number imol (same as
).
Where imol is an integer number
save previous colour map for molecule number imol
Where imol is an integer number
restore previous colour map for molecule number imol
Where t is an integer number
set the state of immediate map upate on map drag.
By default, it is on (t=1). On slower computers it might be better to set t=0.
Where:
- f1 is a number
- f2 is a number
- f3 is a number
set the colour of the last (highest molecule number) map
Where:
- imol is an integer number
- red is a number
- green is a number
- blue is a number
set the colour of the imolth map
Where f is a number
set the sigma step of the last map to f sigma
Where:
- f is a number
- state is an integer number
- imol is an integer number
set the contour level step
set the contour level step of molecule number imol to f and variable state (setting state to 0 turns off contouring by sigma level)
Where imol is an integer number
return the resolution of the data for molecule number imol. Return negative number on error, otherwise resolution in A (eg. 2.0)
Where imol is an integer number
return the resolution set in the header of the model/coordinates file. If this number is not available, return a number less than 0.
Where:
- imol is an integer number
- filename is a string
export (write to disk) the map of molecule number imol to filename.
Return 0 on failure, 1 on success.
Where:
- imol1 is an integer number
- imol2 is an integer number
- map_scale is a number
make a difference map, taking map_scale * imap2 from imap1, on the grid of imap1. Return the new molecule number. Return -1 on failure.
Where val is a number
set the contour scroll step (in absolute e/A3) for 2Fo-Fc-style maps to val
The is only activated when scrolling by sigma is turned off
Where val is a number
set the contour scroll step for difference map (in absolute e/A3) to val
The is only activated when scrolling by sigma is turned off
Where r is a number
set the map sampling rate (default 1.5)
Set to something like 2.0 or 2.5 for more finely sampled maps. Useful for baton-building low resolution maps.
Where is_increment is an integer number
change the contour level of the current map by a step
if is_increment=1 the contour level is increased. If is_increment=0 the map contour level is decreased.
Where level is a number
set the contour level of the map with the highest molecule number to level
Where n_sigma is a number
set the contour level of the map with the highest molecule number to n_sigma sigma
Where i is an integer number
create a lower limit to the "Fo-Fc-style" map contour level changing
(default 1 on)
Where i is an integer number
create a lower limit to the "2Fo-Fc-style" map contour level changing
(default 1 on)
Where f is a number
set the actual map level changing limit
(default 0.0)
Where f is a number
set the actual difference map level changing limit
(default 0.0)
Where f is a number
set the scale factor for the Residue Density fit analysis
Where w is an integer number
draw the lines of the chickenwire density in width w
Where:
- mtz_file_name is a string
- f_col is a string
- phi_col is a string
- weight is a string
- use_weights is an integer number
- is_diff_map is an integer number
make a map from an mtz file (simple interface)
given mtz file mtz_file_name and F column f_col and phases column phi_col and optional weight column weight_col (pass use_weights=0 if weights are not to be used). Also mark the map as a difference map (is_diff_map=1) or not (is_diff_map=0) because they are handled differently inside coot.
Returns: -1 on error, else return the new molecule number
Where:
- mtz_file_name is a string
- a is a string
- b is a string
- weight is a string
- use_weights is an integer number
- is_diff_map is an integer number
- have_refmac_params is an integer number
- fobs_col is a string
- sigfobs_col is a string
- r_free_col is a string
- sensible_f_free_col is an integer number
as the above function, execpt set refmac parameters too
pass along the refmac column labels for storage (not used in the creation of the map)
Returns: -1 on error, else return imol
Where:
- mtz_file_name is a string
- a is a string
- b is a string
- weight is a string
- use_weights is an integer number
- is_diff_map is an integer number
- have_refmac_params is an integer number
- fobs_col is a string
- sigfobs_col is a string
- r_free_col is a string
- sensible_f_free_col is an integer number
- is_anomalous is an integer number
- use_reso_limits is an integer number
- low_reso_limit is a number
- high_reso_lim is a number
as the above function, except set expert options too.
Where:
- mtz_file_name is a string
- f_col is a string
- phi_col is a string
- weight_col is a string
- use_weights is an integer number
does the mtz file have the columms that we want it to have?
Where filename is a string
read MTZ file filename and from it try to make maps
Useful for reading the output of refmac. The default labels (FWT/PHWT and DELFWT/PHDELFWT) can be changed using ...[something]
Returns: the molecule number for the new map
Where i is an integer number
set the flag to do a difference map (too) on auto-read MTZ
return the flag to do a difference map (too) on auto-read MTZ
Returns: 0 means no, 1 means yes.
Where:
- fwt is a string
- phwt is a string
- is_for_diff_map_flag is an integer number
set the expected MTZ columns for Auto-reading MTZ file.
Not every program uses the default refmac labels ("FWT"/"PHWT") for its MTZ file. Here we can tell coot to expect other labels so that coot can "Auto-open" such MTZ files.
e.g. (set-auto-read-column-labels "2FOFCWT" "PH2FOFCWT" 0)
Where f is a number
set the extent of the box/radius of electron density contours
Where f is a number
another (old) way of setting the radius of the map
Where str is a string
Give me this nice message str when I start coot.
Where istate is an integer number
not everone likes coot's esoteric depth cueing system
Pass an argument istate=1 to turn it off
(this function is currently disabled).
native depth cueing system
return the state of the esoteric depth cueing flag
Where i is an integer number
not everone lies coot's default difference map colouring.
Pass an argument i=1 to swap the difference map colouring so that red is positve and green is negative.
Where imol is an integer number
post-hoc set the map of molecule number imol to be a difference map
Returns: success status, 0 -> failure (imol does not have a map)
Add another contour level for the last added map.
Currently, the map must have been generated from an MTZ file.
Returns: the molecule number of the new molecule or -1 on failure
Where imap is an integer number
Add another contour level for the given map.
Currently, the map must have been generated from an MTZ file.
Returns: the molecule number of the new molecule or -1 on failure
return the scale factor for the Residue Density fit analysis
Where:
- imol is an integer number
- x is a number
- y is a number
- z is a number
return the density at the given point for the given map. Return 0 for bad imol
Where imol_map is an integer number
return the mtz file that was use to generate the map
return 0 when there is no mtz file associated with that map (it was generated from a CCP4 map file say).
Where imol_map is an integer number
return the FP column in the file that was use to generate the map
return 0 when there is no mtz file associated with that map (it was generated from a CCP4 map file say).
Where imol_map is an integer number
return the phases column in mtz file that was use to generate the map
return 0 when there is no mtz file associated with that map (it was generated from a CCP4 map file say).
Where imol_map is an integer number
return the weight column in the mtz file that was use to generate the map
return 0 when there is no mtz file associated with that map (it was generated from a CCP4 map file say) or no weights were used.
Where imol_map is an integer number
return flag for whether weights were used that was use to generate the map
return 0 when no weights were used or there is no mtz file associated with that map.
Where:
- imol is an integer number
- file_name is a string
write molecule number imol as a PDB to file file_name
Where:
- imol is an integer number
- chainid is a string
- resno_start is an integer number
- resno_end is an integer number
- filename is a string
write molecule number imol's residue range as a PDB to file file_name