Outline Procedures for Data Integration using SAINT

(March, 2007)

The following is an outline procedure for integrating and reducing the data that was collected on the SMART or APEX systems.  It presumes that you have read the for help file PC Terminal Operations.  In the procedures, as elsewhere in these help files, your input will be in bold-faced type.  As usual, user-determined filenames and such, will be references in bold-faced italics.

All these operations take place on the Linux PC, known as "ChexCSD".  The Linux operating system is case-sensitive, where commands are indicated in the procedures, they should be typed as written (usually in lower case).  To Linux "runbob" and "RunBOB" are very different things.

Procedure for Integrating and Reducing Data

1) Login to ChexCSD and prepare files for integration

On one of the terminals, login to ChexCSD (RMB on the blackbackground and select ChexCSD from the menu)
Username: smartuse or apexuser, as appropriate depending on which instrument you collected data on.
This is important, the two instruments have different characteristics which the software needs to be aware of.

The password, if you do not already know it, can be obtained from Dr. Hollander.

you will now be at a prompt that looks like:
smartuse:/home/smartuse # %       or
apexuser:/home/apexuser # % (the '#' will be a number, we will use <username> to specify the directory from here on)

<username> cd mydir  This should be the same directory name that you used when you transferring your data from the Pentium computers in 32A Lewis using goftp.

<username/mydir> presaint   This command copies the default information file for the detector to your working directory.

<username/mydir> saint   This command starts the SAINT program.  SAINT puts up a list of possible operations.  We are interested in the integration package.  Should you need to utilize the additional operations, Dr. Hollander will assist you.  They are not covered by this primer.
All input in SAINT is terminated with a <CR>, despite the menu-like appearance of the interface.

2) Data integration

SAINT> in   (for integration)  SAINT is a complex and very competent program.  We are only going to do one thing, integrate the reflections.  Selecting in gives us a display with a number of pieces of information (see below) and letters and symbols in [B]rackets.  Typing the letter or symbol in brackets allows us to change the information in that section using the line at the bottom of the screen.  This line wraps around automatically and only accepts the information when you press <Enter> or <CR>.  Most of the defaults are fine as they are, we will normally only change the following:

Command: t -- Input a new, informative, title to be written to your output files.

Command: i -- Type the filename of the first frame of each sequence of frames. This will be the same as what you told SMART to write out, plus the sequence numbers and the frame numbers. In the standard mode you will start with the .001 frame of each series, so you only have to specify it once; the program will assume that each sequence starts with the same number. Example: filename0.001,filename1,filename2,filename3<cr>. The line will wrap at some point during your typing. Don't worry about it. Note that the filenames appear in the display as soon as you finish the entry. If you made an error, type i again and re-enter.

Command: r -- Enter the filenames which will accept the raw output data from integration. It is recommended, for the sake of memory, that you use the same name as your frame files. An exception would be if you were doing a second integration with different parameters. Then you will want to choose a different name. The extension "raw" seems to be required for proper processing. Again, you only need to enter the extension once and ignore the wraparound of the line, . Example: filename0.raw,filename1,filename2,filename3<cr>.

Command: p -- Enter the point group/Laue group symbol and the Lattice-type which corresponds to your data. See Appendix to this help file. Don't worry about it over much if it turns out to be wrong. The only thing this information is used for is to generate statistics on averaging of data. There is no need to re-run SAINT just because the point group was incorrect.

If you have a centered cell, and specify the centering it means that SAINT does not try to integrate the systematic absences due to the cell-centering. This improves things like spot-shape statistics and reduces the size of the raw files by at least a factor of two. You should do this if you know you have a centered cell. If you discover that you have a centered cell, SAINT should be re-run with the lattice type correctly input.

The lattice type only needs to be changed if your cell is not a primitive cell.

NO DATA ARE AVERAGED IN SAINT. Each reflection found is separately written to the filename#.raw and filenamem.raw data files.

Command: s -- Normally you will not change the box size parameters unless you know that your spots were large, or if you have a large unit cell and want to reduce the size of the integration box. Once you have started integration you will have a chance to look at the spot size and decide if you want to change box size. The box X, Y and Z sizes are all preset, you can if you wish change these to smaller or larger values.  The default is a box size of 1.6 x 1.6 x 1.0 degrees. 

Command: u -- This value is the highest resolution of your data set.  Due to the configurations of the two detectors, the SMART has a default resolution of 0.85 A and the APEX defaults to 0.80 A.  If your high angle data are a lower resolution than either of these, we recommend that you reduce the resolution limit to a more reasonable value.  The final merged listing output from SAINT can help you decide if you need to reduce the resolution of your integrated data, or it can be applied in SADABS or XPREP -- we recommened a cut-off where more than 75% of the data are less than 2 sigma(I) after application of SADABS.  This is strictly optional.

Command: y -- If you suspect serious crystal motion, change this value from 0 to 200, or even 100. This will slow down processing a little, but not that much. Not useful unless you actually have some motion.

Command: g -- To set the global refinement of the unit cell based on all strong data observed within the data set.  The point group type should be set to match your crystal system.  NOTE: there are three options for a monoclinic lattice.  In general you should select 3MB, (Monoclinic, B-unique) because this is the standard setting for a monoclinic lattice).

Command: ! -- Typing the exclamation point (or "bang") causes the program to start data processing. The output will start to scroll up your screen. At any point if you want to interrupt the scrolling and look at your output, you can type <ctrl/S>, this stops the output (and holds up processing), but allows you to scroll your window up and down to examine the results. Hitting <ctrl/Q> will unlock the terminal and allow output to continue.

After integrating the data from all frame sets the program will merge the raw data into a data file called filnamem.raw and a corresponding printable output file, filnamem._ls. Print the latter with the command print132 filenamem._ls

If you feel you have to interrupt the program before it finishes, the only way is with <ctrl/C>, which crashes it. This leaves various files in an unfinished state, specifically the *.raw files and the *._ls printable output files. If you have stopped the program, you will want to delete those files before restarting SAINT.

TIPS ON INTERPRETATION OF THE OUTPUT

Return to the Main Help Pages



APPENDIX

Laue and point group information

Point Group Settings

(back to integration instructions)

NAME(s) DESCRIPTION

1-, -1 ................................. Laue group 1 (bar); triclinic system

2/mC ................................... Laue group 2/m, monoclinic C-unique ....NOTE **

2/mB ................................... Laue group 2/m, monoclinic B-unique ....NOTE **

2/mA ................................... Laue group 2/m, monoclinic A-unique ....NOTE **

** NOTE--The letters A, B, and C here indicate the unique axis (ie. the axis that is perpendicular to mirror plane).
They do not indicate a centering condition. Centering is dealt with elsewhere.

4/m ...................................... Laue group 4/m, low-symmetry tetragonal

3-H, -3H ........................... Laue group 3(bar), low-symmetry trigonal, (hexagonal cell -- a=b≠c, gamma=120.)

3-R, -3R ............................ 3-bar, low-sym. trigonal, (rhombohedral primitive cell setting -- a=b=c.)

6/m ...................................... Laue group 6/m, low-symmetry hexagonal

m3-, m-3 ........................... Laue group m3(bar), low-symmetry cubic

mmm ................................... Laue group mmm, orthorhombic system

4/mmm ................................ Laue group 4/mmm, high-symmetry tetragonal

6/mmm ................................ Laue group 6/mmm, high-symmetry hexagonal

3-m1, -3m1, 3m1 ............... Laue group 3(bar)m1, high-symmetry trigonal

3-1m, -31m, 31m ............... Laue group 3(bar)1 m, high-symmetry trigonal

3-m, -3m,............................ Laue group 3(bar)m, high-symmetry trigonal, (rhombohedral primitive cell setting)

m3-m, m-3m, m3m ............ Laue group m3(bar)m, high-symmetry cubic

(back to integration instructions)

Lattice types

Enter the number corresponding to the lattice type to be used for exclusion of systematic absences in the output. (return to integration instructions)

0 P -- Primitive

1 A -- A-centered

2 B -- B-centered

3 C -- C-centered

4 F -- Face-centered

5 I -- Body-centered

6 RO -- Rhombohedral obverse

7 RR -- Rhombohedral reverse

At this stage, there is no reason you should need to know (and hence no need to input) more specific space-group symmetry. XPREP needs the intensities of possible space-group absences in order to determine or verify the space group. (return to integration instructions)

Return to the Main Help Pages