. DEMONSTRATION~NOTES~FOR~LAUE~DATA~PROCESSING ============================================~ . Marjorie Harding and John Campbell December 1992 @ SECTION~1 . INTRODUCTION~ ============~ . A fairly full account of the principles used, in the majority of the programs in the Laue software suite, is given in Helliwell et. al (ref 1.) which should be read in conjunction with the instructions for the individual programs. It is also important to recognise that some of the basic framework is derived from the Cambridge/Imperial College suite of programs for the processing of monochromatic oscillation photographs of protein crystals (refs. 2,3). Thus the first objective is to find the crystal orientation and predict a set of spot positions. Then refinement is carried out for the crystal orientation angles, camera constants and all the parameters necessary to make the predictions accurate to, say, 0.05 mm. This is followed by integration of the optical density in a box around each predicted spot position and subtraction of the background. The demonstration examples are designed to help the beginner in Laue data processing to get experience with data which can be processed tolerably well by each of the main programs involved. Those already familiar with the processing of data using monochromatic radiation must 'think again' as the wavelength is now a variable! It is necessary to have some understanding of the coordinate systems in use and some details of these are given in the final section of these notes. The current version of the demonstration is a modified version of that prepared for an EEC sponsored Laue workshop held at the Daresbury Laboratory in October 1989. Image plate data processing is now included. a)~ Examples^provided~ The three examples used are: 1) PF~ which is proflavin hemisulphate, (C13 H12 N3)2.SO4.(3.5)H2O, monoclinic, space group P21/C with 8 formula units per unit cell; excellent quality crystals were provided by Dr S. Neidle and have been used for many test photographs. (Structure determined, Jones and Neidle, 1975, Acta Cryst. B31~ 1324) This example works very straightforwardly at all stages. Because dmin is quite small (< 1 Angstrom) there are as many spots on the film as might be found for a small protein structure. 2) LYS~ which is HEWL tetragonal Lysozyme. The Laue data collected on a single image plate (MAR Research system). Again this data may be processed in a straightforward manner. Diffractometer data for reference were made available by Dr J. Dewan. 2) ZN~ which is 4 zinc insulin, space group R3, 18 molecules (of molecular weight about 6000) in the hexagonal unit cell given. The structure is known and well refined (eg Bentley, Dodson et al., Nature, 1976, 261~, 166-8). In the current demonstration, this is only used to illustrate the processing of 'difference' laue data. Image plots of the PF top film and the LYS image are reproduced at the end of this section. b)~ Setting~Up^ The user should set up a directory/sub-directory for running the demonstration. Apart from the film and image-plate image files, all other files required should be copied to the user's area from the supplied test data area. The command 'lget' has been set up to facilitate such copying. The command is followed by the names of one or more files which are to be copied (maximum of 8 names on the Vax). These filename specifications may contain wild cards (will need to be enclosed in " " if under Unix). Some of the files used by the Laue programs are in machine dependent binary format and such files are supplied in ascii format and will be translated by the 'lget' command to the corrsponding binary files (more details below). E.g. (for testing the program Lauescale) lget lys_ls.ctl lys_ls.geasc lys_ls.mtzasc This will copy the three named files to the user's area and will translate the file lys_ls.geasc to lys_ls.ge1 and lys_ls.ge2 and also translate the file lys_ls.mtzasc to lys_ls.mtz. The film and image-plate files need not be copied but are handled as follows: Under^Unix Soft links are made to the image files from the user's area. The files may then be referenced from the user's area as if they were present in that area. This may be accomplished by using the supplied command 'linkimages'. This will make links to all the available image files. Under^VMS The logical name 'lfilm' has been set up to point to the directory containing the test data image files. This is used when accessing a required image file. e.g. lfilm:pf1a.dat Note: On a Unix system the test data files are held in the directory $CCP4_MASTER/laue/testdata and under VMS they are stored in the directory CCP4_ROOT:[XTAL.LAUE.TESTDATA]. c)~ Some~Notes~on~the~Supplied~Files^ Film/Image-plate^images The images supplied are as follows: a) Proflavin: The 'a' and 'b' films of a 6 film pack pf1a.dat and pf1b.dat. b) Lysozyme: A single MAR image-plate image lys.i2 (lys_swap.i2) Note: For the image-plate data, the data items are unsigned two byte integers. Most Unix based machines are big-endian. For a little-endian Unix machine, the byte swapped image lys_swap.i2 should be used instead of lys.i2. (On the Vax, lys.i2 has the correct byte order for the Vax, being the opposite of that in lys.i2 supplied on a Unix system) Laue^.ge1/.ge2^files These are the files which contain the lists of the predicted reflections and store the integrated intensities. As used by the Laue programs they are in a binary format. As such binary data is machine dependent, the files are supplied in an ascii format which needs to be translated to the required binary format using the supplied program 'exchge' with the option 'translate from ascii (f)'. If the supplied ascii files required for a test are copied to the user's directory using the supplied 'lget' command then the .geasc files will be automatically translated to the corresponding .ge1/.ge2 files. MTZ^files These are also files which contain binary data. Again they are supplied in an ascii format which needs to be converted to the binary MTZ format for use using the supplied program 'na4tomtz'. The program may be run by typing the command 'laue na4tomtz'. If the supplied ascii files required for a test are copied to the user's directory using the supplied 'lget' command then the .mtzasc files will be automatically translated to the corresponding .mtz files. d)~ Documents^~ In addition to this document, you should refer to the following documents. 'Introduction to the Laue software suite'. This contains a flow chart showing the normal sequence of programs used for processing Laue data. (Distributed with the Laue Software Suite) User documentation for the major programs which will be used. (Distributed with the Laue Software Suite) A copy of the paper 'The Recording and Analysis of Synchrotron X-radiation Laue Diffraction Photographs' (See reference 1) e)~ References^~ 1) Helliwell, J.R., Habash, J., Cruickshank, D.W.J., Harding, M.M, Greenhough, T.J., Campbell, J.W., Clifton, I.J., Elder, M., Machin, P.A., Papiz, M.Z. and Zurek, S. (1989) J. Appl. Cryst, 22~ part 5. 2) Arndt, U.W. & Wonacott, A.J., (1977). "The Rotation Method in Crystallography" Amsterdam: Netherlands. 3) Wonacott, A.J. (1980) Notes on a suite of programs for processing oscillation camera film data. (MOSFLM documentation) @ REPLACE~THIS~PAGE~WITH~PROFLAVIN~FILM~IMAGE~PICTURE Use 'filmps' program with pf1a.dat All defaults except: file name short title: Proflavin 'A' film no further annotation @ REPLACE~THIS~PAGE~WITH~LYSOZYME~IMAGE~PICTURE Use 'filmps' program with lys.i2 All defaults except: image plate file name scaling thresholds: 0 1000 short title: Lysozyme Image-plate Image no further annotation @ SECTION~2 . FINDING~THE~CRYSTAL~ORIENTATION~-~NEWLAUE =========================================~ . Introduction^~ The program NEWLAUE has two uses: a) for solving for the crystal orientation when the unit cell and crystal to film distance are known, but not the orientation angles PX (PhiX), PY (PhiY), PZ (PhiZ). This uses a file of spot positions made with the program SPOTIN which uses functions from the program GENLAUE and works in a similar way. b) for predicting Laue patterns given the unit cell, orientation of the crystal with respect to the beam and spindle, crystal to film distance, and the 'soft' limits (lambda-min, lambda-max and dmin). The program is very useful for learning about the effects of each of these parameters on the Laue pattern produced. You are reminded that the orientation angles are defined in the same way as in the oscillation film processing programs with the axes X parallel to the beam and Z parallel to the spindle. Example^1^-^proflavin^hemisulphate^(PF)~ Test data files required: pf_nl.dmp (file of spot positions) Crystal data: a=12.703, b=19.940, c=21.487, beta=92.29 degrees. Monoclinic, Primitive Cell Crystal to film distance (DISTAN): 61.0 mm Camera type is the default (C1) To run NEWLAUE, type: laue newlaue Input the items of data as requested, noting the following: . WMIN 0.3A, WMAX 2.0A are reasonable DMIN 1.4A is useful although the true dmin is < 1A RADIUS is 59mm SYSTEM ... Carriage return gives Help information. ORIENT ... Choose option for 'misorientation angles' as used in oscillation film processing. Choose a* parallel to spindle and b* parallel to beam in this trial for consistency with our files. The camera type is the default (C1). DATSAV ... will write a .gen file with the current parameter values. This may be used by the program GENLAUE or re-input to NEWLAUE using the DATGET command. e.g. pf1_sav.gen . a) Solving^the^Orientation Leave PX, PY, PZ, SPINDL = 0. START is essential to calculate the reciprocal lattice points. PLOT is pretty but not essential at this stage. AUTO initiates the auto-indexing option. Note the following: . No indices available Input spot positions from file pf_nl.dmp Error in positions < 1.0 mm Refine angles? Yes, and repeatedly till convergence. . This should give a solution with an rms deviation of < 0.3 mm. Keep a record of it. Reset PX, PY, PZ with this solution. b) Checking^the^Orientation Perform the following command sequence: . Check that PX, PY and PZ now set as found; if not reset these. START to calculate reciprocal lattice etc. PLOT and check that displayed simulation agrees with photo. . A plot of this NEWLAUE solution is shown at the end of this section. The spots selected for auto-indexing are numbered on the plot. The DATSAV command may be used to save the parameters for use in the next stage of processing - suggested name pf_genl.gen (use the current .gen file specific parameter values). c) Additional^Explorations To explore the properties of Laue patterns, try any or all of the following, or other variations: . Change PX, PY or PZ by 3 or 6 degrees Change SPINDL by 6 degrees Change WMIN to 0.6A Change DMIN to 2.0A or to 0.9A Index some spots using INDEX option . Example^2^-^Lysozyme^(LYS)~ Test data files required: lys_nl.dmp (file of spot positions) Crystal data: a=79.19, c=38.02 (Tetragonal, Primitive) WMIN = 0.4, WMAX = 2.2 Crystal to film distance (DISTAN): 199.97 mm Suggested DMIN about 2.3A. RADIUS = 90.0 Choose a* parallel to spindle, b* parallel to beam Follow the general procedure learnt from example 1. Note the comments below on Auto-indexing. The DATSAV command may be used to save the parameters for use in the next stage of processing - suggested name lys_genl.gen (use the current .gen file specific parameter values). A plot of this NEWLAUE solution is shown at the end of this section. The spots selected for auto-indexing are numbered on the plot. Comments^on^Auto-indexing~ The procedure depends on converting nodal spot positions to angles between simple reciprocal lattice vectors, and then matching them with calculated ones. 4 good nodals, all with low enough indices, are essential; more may or may not be helpful. The crystal to film distance must be correct and accurate and the shorter it is the better. When taking photos it may be useful to take a few extras at different spindle angles to increase the chance that one can be solved. Hints, comments for difficult cases: a) Raise maximum h**2 + k**2 +l**2 value; use higher values for centred lattices. b) Raise error level slightly. c) Possibly vary crystal to film distance slightly (+ or - 0.5 mm) d) Success depends on 4-8 good nodals with well measured positions. Many more nodals just make the solution slower. Always try refining angles even if the rms looks very high. Comments^on^Properties^of^Laue~patterns It is useful to note the following: a) One spot may be the result of the superposition of several reflections, 'harmonics'. b) Short wavelength reflections occur predominantly near the film centre and long wavelength reflections predominantly nearer the outer edge, but they are intermingled. c) Increasing the dmin value makes the pattern sparser; it does not necessarily remove the outer (high theta) parts of it. d) Nodal spots, roughly speaking, are those which occur at the intersection of several festoons, and are well separated from their neighbours. They have simple indices and are often multiples e.g. 1 1 0 (with 2 2 0, 3 3 0 ...) 1 2 -1 (with 2 4 -2, 3 6 -3 ...) In subsequent programs a spot is defined as a nodal if none of its indices (of the lowest harmonic) is greater than a specified 'nodal index'. @ REPLACE~THIS~PAGE~WITH~POSTSCRIPT~PRINT~OF~pf_sp.ps @ REPLACE~THIS~PAGE~WITH~POSTSCRIPT~PRINT~OF~lys_sp.ps @ SECTION~3 . REFINING~THE~ORIENTATION~AND~GENERATING~THE~.GE~FILES =====================================================~ . Introduction^~ The program GENLAUE is used to refine the crystal orientation and produce files containing the indices and positions of spots to be integrated. It has many options an possibilities and the user should look at the flowchart (given in the program documentation and reproduced below) to help in its use. Refinement is likely to require quite a number of cycles, perhaps 5 - 15, with gradually changing conditions. It is usually desirable to refine all the 'hard' parameters - orientation angles, crystal to film distance etc. and, if appropriate, the cell dimensions - then to consider and optimise the soft limits, lambda-min, lambda-max and dmin and then to re-enter GENLAUE and write the files of predicted spot positions. A relatively high value of dmin can be used in GENLAUE, to save time, when the crystal parameters are being refined but, for the integration of the intensities, a realistic value of dmin must be used in order to assign multiple spots correctly - i.e. you cannot truncate to 'low resolution' data. The program will create the files name.ge1 and name.ge2 (name selected by the user) which contain the predicted positions for all the reflections in one film pack and space for the measured I and sig(I) values on all six films, as well as header information. The .ge1 file is for single and double Laue spots and the .ge2 file is for higher multiple spots. The same format is used for the image plate data though only a single plate is present. Example^1^-^proflavin^hemisulphate^(PF)~ Test data files required: Film image ./pf1a.dat (Unix) (After using 'linkimages') or lfilm:pf1a.dat (Vax) pf_in.gen (parameters file) Carry out the following steps: 1) Copy^and^examine^the^.gen^file Copy the file pf_in.gen from the test data area to your own directory and examine it. PX, PY and PZ should correspond to those found from the autoindexing or be a symmetry equivalent by angles such as 180-PX etc. Alternatively you may rename your output file from NEWLAUE (e.g. pf_genl.gen) as pf_in.gen if you are satisfied with it. 2) Run^GENLAUE^to^refine^the^parameters The program is run by typing the command: laue genlaue For the input file specify pf_in without an extension assuming that you have prepared a file pf_in.gen as just described. Specify 'f' for film data when requested. A flow diagram is provided at the end of this section showing the various routes through the program. Choose a fairly high 'dmin' (e.g. 1.7) for the first run and use default values when in doubt. Choose the refinement option. Choose to predict all~ spots and to match on a threshold plot. You should save an image file in case you need to repeat the process. **Warning**~ When printing a threshold plot there may be some delay before plotting starts on some systems so be prepared to wait. This run should show convincingly that the patterns match approximately and should allow cursor input of pairs of observed and predicted spots to improve the match. Carry on into refinement and explore its possibilities - you may refine just px, py, pz or more parameters. When the predicted and observed patterns are matching within approximately 0.5 to 1.0 mm, it is safe to go over to refining 60-100 nodal spot predictions (choose an appropriate nodal index) against centre of gravity positions found by the program from the digitised image file (i.e. at this stage there is no need to display the plot or use the cursor). Bear in mind that the actual spot size is 0.3-0.4 mm, i.e. 6-8 rasters. It should be possible to refine to get a rms value of <= 0.05 (max 0.10) mm. 3) Writing^the^.ge1/.ge2^files When you are satisfied with your refinement, use the option to write the .ge1/.ge2 files for the integration.program. Name the .ge files pf1 (-> pf1.gen, pf1.ge1, pf1.ge2) When asked the question about crystal to film distances, do~not accept the refined distances. Give the following spacings: 0.20 0.40 0.40 0.40 0.40 (mm) as the film pack is made up as follows: Fa Fb (F) Fc (F) Fd (F) Fe (F) Ff Note that, between stages 2 and 3, it would normally be necessary to experiment further to find good values of lambda-max lambda-min and dmin. Example^2^-^Lysozyme^(LYS)~ Test data files required: Image ./lys.i2 (Unix) (After using 'linkimages') or lfilm:lys.i2 (Vax) lys_in.gen (parameters file) Again you may use the output file from NEWLAUE (e.g. lys_genl.gen) if you are satisfied with it by renaming it as lys_in.gen instead of using the supplied version of lys_in.gen. The procedure is very similar to that described for the first example bearing in mind the following: This is image plate data Use a spot size of 0.6 mm, box size 0.6 mm. You can use a nodal index of say 5 to search for spots (and a threshold of 100 above background for finding spot positions) without needing to match spots on a threshold plot. Refined to rms of around 0.12 mm. Name the o/p .ge files lys1 (-> lys1.gen, lys1.ge1, lys1.ge2) @ GENLAUE~FLOW~DIAGRAM . START~ | Input .gen file |<----------------------------------------------- Calculate spot positions | (Questions about overlaps, nodals etc.) | | | (Display predicted pattern) | | | --<--[NO]--Refinement?--[YES]---- | | | | | Spot posns. | (Refine another film -<-[.dmpfile]--from ?---[image file]-- | in pack? (Crystal to | | | film distance only)) | Fetch digitised image. | | | Find fiducials and | Calculate all spot | background. | positions, write to | | | .ge1, .ge2 files. | Display film image as | | | --<---[NO]------threshold plot ? | Questions about C-F | | | | distances. | | [YES] | | | Find centres of | | Rewrite .gen file w. | gravity of density (Save image as .img | additional parameters | around predicted file for quicker | | | positions display next time) | END~ | | | | | | Pick spot positions | | | with the cursor | | | | | | ------------------------- | | | | | Save spot positions | | in .dmp file | |--------------------------| | Display difference map | (obs-calc positions) | |<---------------- | Select parameters | | for refinement | | | | | Refine | | | | | (Display difference map) | | ( ) indicates optional choices | | | Further refine?--[YES]-- | | | [NO] | |---------------------- . @ SECTION~4 . INTEGRATION~USING~THE~PROGRAM~'INTLAUE' =======================================~ . Introduction^~ The examples use the profile fitting option in INTLAUE which gives better intensities for the weaker reflections than simple 'box' integration. They do not attempt to integrate spots closer than the spatial overlap limit as set in GENLAUE. Laue diffraction spots from good quality crystals will normally be circular in all parts of the film and comparable in size to the collimator used; normally in poorer quality crystals or radiation damaged crystals, the spots are radially^ elongated. For details of the program INTLAUE, see the program documentation and see also the flow diagram at the end of this section. Proflavin^hemisulphate^example~ . Test data files required: Film images ./pf1a.dat & ./pf1b.dat (Unix) (After using 'linkimages') or lfilm:pf1a.dat & lfilm:pf1b.dat (Vax) . pf_intl.gen (.gen file) pf_intl.ctl (Unix) or pf_intl.ctlvax (Vax) (Intlaue control file) pf_intl.geasc -> pf_intl.ge1, pf_intl.ge1 (.ge1/.ge2 files) Note: When pf_intl.geasc is copied from the test data area using the 'lget' command, the files pf_intl.ge1 and pf_intl.ge2 will be generated automatically. As an alternative, if you are happy with your GENLAUE run you may copy your pf1.gen to pf_intl.gen, pf1.ge1 to pf_intl.ge1 and pf1.ge2 to pf_intl.ge2. Lysozyme^example~ . Test data files required: Image ./lys.i2 (Unix) (after using 'linkimages') or lfilm:lys.i2 (Vax) . lys_intl.gen (.gen file) lys_intl.ctl (Unix) or lys_intl.ctlvax (Vax) (Intlaue control file) lys_intl.geasc -> lys_intl.ge1, lys_intl.ge1 (.ge1/.ge2 files) Note: When lys_intl.geasc is copied from the test data area using the 'lget' command, the files lys_intl.ge1 and lys_intl.ge2 will be generated automatically. As an alternative, if you are happy with your GENLAUE run you may copy your lys1.gen to lys_intl.gen, lys1.ge1 to lys_intl.ge1 and lys1.ge2 to lys_intl.ge2. Suggested^trials~ Note that whenever INTLAUE is run, the .ge1/.ge2 files will be updated. Therefore if you wish to re-run the program, you should make fresh copies of the required .ge1/.ge2 files (either copies of the files from your GENLAUE run or copies obtained again from the test data area). Try starting an interactive run using the Proflavin data as follows: Run INTLAUE by typing the command 'laue intlaue' (Use the default 'terminal' as the reply to the prompt DATA: ) You may use default replies in answer to most of the questions except~ . Inner bin radius 200 (in rasters, edge of film is 1800) Outer bin radius 750 Do not use variable radial masking Approximate number of refinement nodals 200, 400 Box size 13 13 5 1 1 (probably - see printed profiles as check) . *Warning*: These last parameters are in fixed 5I5 format. Leave the inner refinement threshold at 30 The 'inner refinement' should give an rms approximately <= 2 The 'outer refinement' should give an rms approximately <= 2 to 3 You should probly terminate the process after these refinrments when the following prompt is output: Terminate process now (Y/N)? The program would go on at this stage to collect density, form profiles etc. You may then re-submit the integration as a batch (Vax) or background job (Unix). You are probably best to start with fresh copies of the .ge1/.ge2 files. Run the program using the command 'laue intlaue' DATA file is pf_intl.ctl (Unix) or pf_intl.ctlvax (Vax) If you do not wish to wait, you may look at a sample output log from the test data area (pf_intl.log) The LYS example may be run in batch/background mode in an analogous manner. The DATA file is lys_intl.ctl (Unix) or lys_intl.ctlvax (Vax). There is a sample log file, lys_intl.log, in the test data area. Remember that this is image-plate data. @ INTLAUE~FLOW~DIAGRAM . START~ | Input files .ge1, .ge2, .gen digitised film image file | | - - - - - - - | | | | | 'Inner' refinement of parameters | | | - -<- - - - - | |<----------------------------------- | | - - - - - - - | | | | | | | 'Outer' refinement of parameters | | | | | - -<- - - - - | | | | Write improved parameters to .ge1, .ge2 files | | | Collect density | | | Form profiles - A film only, normally | | | Fit profiles, write intensities to .ge1, .ge2 files | | | Output summary, statistics | | | Any more films in pack? -----[YES]-------------- (end with F) | [NO] | END~ . - - - - Interactive use only, not batch @ SECTION~5 . FILM~PACK~SCALING~USING~THE~PROGRAM~'AFSCALE' =============================================~ . Introduction^~ This program needs to find and apply film factors between successive films of one pack; each film factor is a wavelength dependent function and is modelled as exp(alpha*lambda**3) with three different values of alpha (described as Victoreen coefficients; The coefficients k in the expression k*exp(alpha*lambda**3) are now kept equal to 1.0). The three alpha coefficients are for wavelength ranges separated by the discontinuities in the scaling functions at the Silver and Bromine absorption edges, at 0.49 and 0.92 Angstroms, due to the Silver and Bromine in the film. Proflavin^hemisulphate^example~ Test data files required: pf_af.geasc (-> pf_af.ge1, pf_af.ge2) (.ge1/.ge2 files containing integrated intensities for A and B films) pf_af.vc (Victoreen coefficients) Note: When pf_af.geasc is copied from the test data area using the 'lget' command, the files pf_af.ge1 and pf_af.ge2 will be generated automatically. As an alternative, if you are happy with your INTLAUE run you may copy your pf_intl.ge1 to pf_af.ge1 and pf_intl.ge2 to pf_af.ge2. Running^the^program~ The program is run by typing the command: laue afscale After initial input of the intensity data (.ge1) file and the starting Victoreen coefficients file (.vc) the usual procedure for each film pair will involve: . D define - to select reflection pairs to be used and F fit - to evaluate the three alpha coefficients . followed by: Plot, Scale and/or Rsym to assess the results of the scaling The above steps will probably need to be repeated several times with variations in the parameters set by Define. Experiment with the program to get the best coefficients for the film pair A/B. When the film pair A/B (normally all film pairs A/B ... E/F) has been thus fitted, the scaled and merged intensities are output to a file (pf1.afout). The Victoreen coefficients may also be written to a file (pf1.vc). See below for a summary of AFSCALE actions. Summary^of^the^program^actions~ . Input~ Define an input .ge1 file Load~ Load a .vc (approximate starting coefficients) file Define~ Selects reflections to be used | for finding parameters - on the | Repeat these basis of film pairs, exclusions | two till happy on I, sig(I) etc. } for all pairs | of films Fit~ Fits 6 parameters | Plot~ | Scale-R~ } Useful to monitor the goodness of the fit Rsym~ | Write~ Finally write new .VC parameters to file Output~ Finally write scaled and merged intensities to file . @ SECTION~6 . WAVELENGTH~NORMALISATION~USING~THE~PROGRAM~LAUENORM ===================================================~ . The program LAUENORM is used to perform an internal wavelength normalisation of the intensity data. Proflavin^hemisulphate^example Test data files required: pf_ln.ctl (control data file) pf1_ln.afout, pf2_ln.afout, pf3_ln.afout (AFSCALE output files) Note: If the AFSCALE run was satisfactory the output file pf1.afout may be copied to pf1_ln.afout and used in the normalisation run; the other two files will still need to be copied from the test data area. Run the program using the command 'laue lauenorm' . DATA: pf_ln.ctl~ LAUEHKL1: pf1_ln.afout~ LAUEHKL2: pf2_ln.afout~ LAUEHKL3: pf3_ln.afout~ LAUEHKL4: HKLOUT: pf_ln.mtz~ . Run in background (or batch) mode. The output log file can be compared with the example log file pf_ln.log. If required, the program mtzdump can be used to dump the contents of the created MTZ file e.g. pf_ln.mtz. A log file of such a dump with the first 100 reflections is available for comparison as pf_ln.mtzd. Look at the number of overlaps available for the scaling and look at the internal agreement factors obtained. You could try running LAUENORM for one of the individual files for PF. You will need to edit the control file pf_ln.ctl for this. Change the title line and change 'NORMALISE 3' to 'NORMALISE 1'. What causes the problems encountered? Lysozyme^example For image plate data with a single image plate no 'film to film' scaling is needed and there is an option to run LAUENORM using data directly from the integration program INTLAUE. Test data files required: lys_ln.ctl (control data file) lys_ln.geasc (-> lys_ln.ge1, lys_ln.ge2) (.ge1/.ge2 files after INTLAUE) Note: When lys_ln.geasc is copied from the test data area using the 'lget' command, the files lys_ln.ge1 and lys_ln.ge2 will be generated automatically. As an alternative, if the INTLAUE run was satisfactory, the output file lys_intl.ge1 may be copied to lys_ln.ge1 and lys_intl.ge2 to lys_ln.ge2 and used in the normalisation run. Run the program using the command 'laue lauenorm' . DATA: lys_ln.ctl~ LAUEHKL1: lys_ln.ge1~ LAUEHKL2: HKLOUT: lys_ln.mtz~ . Run in background (or batch) mode. The output log file can be compared with the example log file lys_ln.log. If required, the program mtzdump can be used to dump the contents of the created MTZ file e.g. lys_ln.mtz. A log file of such a dump with the first 100 reflections is available for comparison as lys_ln.mtzd. Look at the number of overlaps available for the scaling and look at the internal agreement factors obtained. @ SECTION~7 . WAVELENGTH~NORMALISATION~USING~THE~PROGRAM~LAUESCALE ====================================================~ . The program LAUESCALE is used to perform a wavelength normalisation of the intensity data against a reference set of data. You should compare the normalisation results with those obtained via LAUENORM. LAUESCALE normalises only a single pack at a time. Proflavin^hemisulphate^example Test data files required: pf_ls.ctl (control data file) pf_ls.afout (AFSCALE output file) pf_ls.mtzasc (-> pf_ls.mtz) (Reference data) Note: When lys_ls.mtzasc is copied from the test data area using the 'lget' command, the file lys_ls.mtz will be generated automatically. Also: If the AFSCALE run was satisfactory the output file pf1.afout may be copied to pf1_ls.afout and used in the normalisation run. Run the program using the command 'laue lauescale' . DATA: pf_ls.ctl~ LAUEHKL: pf_ls.afout~ HKLIN: pf_ls.mtz~ HKLOUT: pf_lsout.mtz~ HKLOUT2 (optional): SHELX (optional): DIAGNOSTICS FILE NAME (optional): . Run in background (or batch) mode. The output log file can be compared with the example log file pf_ls.log. If required, the program mtzdump can be used to dump the contents of the created MTZ file e.g. pf_lsout.mtz. A log file of such a dump with the first 100 reflections is available for comparison as pf_lsout.mtzd. Lysozyme^example For image plate data with a single image plate no 'film to film' scaling is needed and there is an option to run LAUESCALE using data directly from the integration program INTLAUE. Test data files required: lys_ls.ctl (control data file) lys_ls.geasc (-> lys_ls.ge1, lys_ls.ge2) (.ge1/.ge2 files after INTLAUE) lys_ls.mtzasc (-> lys_ls.mtz) (Reference data) Note: When lys_ls.geasc is copied from the test data area using the 'lget' command, the files lys_ls.ge1 and lys_ls.ge2 will be generated automatically. Similarly lys_ls.mtz will be generated from lys_ls.mtzasc. As an alternative, if the INTLAUE run was satisfactory, the output file lys_intl.ge1 may be copied to lys_ln.ge1 and lys_intl.ge2 to lys_ln.ge2 and used in the normalisation run. Run the program using the command 'laue lauescale' . DATA: lys_ls.ctl~ LAUEHKL: lys_ls.ge1~ HKLIN: lys_ls.mtz~ HKLOUT1: lys_lsout.mtz~ HKLOUT2 (optional): SHELX (optional): DIAGNOSTICS FILE NAME (optional): . Run in background (or batch) mode. The output log file can be compared with the example log file lys_ls.log. If required, the program mtzdump can be used to dump the contents of the created MTZ file e.g. lys_lsout.mtz. A log file of such a dump with the first 100 reflections is available for comparison as lys_lsout.mtzd. @ SECTION~8 . PROCESSING~OF~DIFFERENCE~DATA~FOR~TIME~RESOLVED~STUDIES =======================================================~ . The program DIFFLAUE is used to prepare a file of difference Laue data scaled to a reference set of data. The program requires two input .ge1 files (for the two sets of laue data measured for the same crystal setting at different times) and an input reference set of data corresponding to the crystal state at time 1. Insulin^example Test data files required: zn_dl.ctl (control data file) zn1_dl.geasc, zn2_dl.geasc (-> zn1_dl.ge1, zn1_dl.ge2, zn2_dl.ge1, zn2_dl.ge2) (.ge1/.ge2 files after INTLAUE) zn_dl.mtzasc (-> zn_dl.mtz) (Reference data) Note: When zn1_dl.geasc and zn2_dl.geasc are copied from the test data area using the 'lget' command, the files zn1_dl.ge1, zn1_dl.ge2 and zn2_dl.ge1, zn2_dl.ge2 will be generated automatically. Similarly zn_dl.mtz will be generated from zn_dl.mtzasc. Run the program using the command 'laue difflaue' . DATA: zn_dl.ctl~ LAUEGE1: zn1_dl.ge1~ LAUEGE2: zn2_dl.ge1~ HKLIN: zn_dl.mtz~ HKLOUT: zn_dlout.mtz~ . Run in background (or batch) mode. The output log file can be compared with the example log file zn_dl.log. If required, the program mtzdump can be used to dump the contents of the created MTZ file e.g. zn_dlout.mtz. A log file of such a dump with the first 100 reflections is available for comparison as zn_dlout.mtzd. @ SECTION~9 . OTHER~PROGRAMS~OR~FACILITIES ============================~ . 1) To make a .dmp file for use in NEWLAUE use the command 'laue spotin' The SPOTIN program contains simplified sections from the program GENLAUE and is run in a similar manner for the specific purpose required. 2) To Unscramble Multiple Spots The program UNSCRAM may be run after AFSCALE. It uses the .ge1 and .ge2 files and the .vc file written by AFSCALE. (Test data files are also available as pf_unsc.geasc (->pf_unsc.ge1, pf_unsc.ge2) and pf_unsc.vc) The program may be run by typing the command: laue unscram 3) To get an overview of the spread and quality of data in the output from AFSCALE or for individual films in .ge1 files use the program initiated via the command 'laue intanal' The data control file pf_ia.ctl is available for input from a .ge1 file. The .ge1/.ge2 files from the INTLAUE run on the proflavin data may be used or alternatively the test data file pf_ia.geasc (-> pf_ia.ge1, pf_ia.ge2) may be used. 4) The program LCHK analyses the distribution of symmetry_unique reflections in a Laue pattern. LCHK is intended to complement NEWLAUE and most of the conventions assumed in NEWLAUE are followed in LCHK. A menu allows for the input of parameter values and for the selection of a number of different commands for providing analyses of the data which would be collected for various combinations of crystal settings. The program is run by typing the command 'laue lchk' A test data file pf_lchk.dat is available if you do not wish to set up your own parameters. @ SECTION~10 . COORDINATE~SYSTEMS ==================~ . Basic~Coordinate~Systems^ The relationship between the 'film' coordinate system (xf, yf) used by the Laue software suite and the 'Laboratory Coordinate System' (X, Y, Z) is shown in the first of the two diagrams at the end of this section. Scanner~Coordinates~for~Films^ The drum co-ordinate system used by the program is defined with Xd along the axis of the drum and Yd around the cylindrical surface of the drum. Co-ordinates are measured in scanner units of 25, 50 or 100 microns. The film is mounted on the drum with the rotation axis along Xd. ie in the orientation shown in the second diagram at the end of this section. The numbering of the fiducials must be as shown in that diagram for the two cameras in use at Daresbury. The direction of drum rotation on Optronics microdensitometers is opposite to that of the Joyce Loebl Scandig, and consequently the direction of Yd is effectively reversed. To allow for this the film must be inverted top-to-bottom when placed on the drum of an Optronics machine. This is the only change required, and the fiducial coordinates should be given relative to the Scandig coordinate frame ( Xd along the drum axis from left to right, Yd around the drum in an anticlockwise direction). Camera~Constants^ These are parameters which relate the position of the fiducial spots to the position and orientation of the film co-ordinate frame (xf, yf). They are defined relative to the centre of fiducial marks 1 and 3 and the line joining fiducials 2 and 3. Thus Xcen = ( x[1] + x[3] )/2 + ccx Ycen = ( y[1] + y[3] )/2 + ccy w = ARCTAN(( y[3] - y[2] )/( x[3] + x[2] )) + ccomega where (ccx, ccy, ccomega) are the camera constants and Xcen,Ycen and w (omega) are three of the parameters which define the transformation between the co-ordinate frames of the film and the drum. In cases where there are no fiducuals (e.g. for image plates), values of Xcen and Ycen are input by the user and w=0. (See also the second diagram at the end of this section) Transformations~^ The basic parameters which define the transformation between film co-ordinates (xf, yf) and scanner co-ordinates (Xd, Yd) (or in general the coordinates in raster units in the image file) are Xcen, Ycen the position of the film centre or direct beam, w the orientation of the film on the drum (ie the angle between Xf and Xd) and K a scale factor (note that K is approximately equal to 1.0 because the crystal-to-film distance is implicit in the film co-ordinates). A further scale parameter 'yscale' is used to define the relative scales along Yd and Xd in the drum co-ordinate frame. This is necessary because X-ray film is double-sided and the film image is at a slightly larger radius than the drum surface when the film is mounted. The effect is to increase the separation between samples along Yd by approximately 0.2%. Thus yscale = 0.998 for film. The transformation is given by Xd = F*[ Xcen + K*( xf*cosw - yf*sinw)] Yd = F*[ Ycen + K*yscale*( xf*sinw + yf*cosw )] where F is the factor which converts from millimetre to raster units. (See also the second diagram at the end of this section) Film~Distortion^ The only kinds of distortion that are likely to occur in the recording of the diffraction pattern on flat films are due either to the film plane not being normal to the X-ray beam or to the film not sitting flat in the cassette. For flat film refinement three additional parameters have been introduced in the transformation to allow for distortion; these are TILT, TWIST and BULGE. TILT and TWIST correct for non-normal film plane which can occur with a poorly aligned or incorrectly located film cassette. In the expressions for the film/drum transformation the scale factor, K is replaced by K', as given by K' = K + TILT*xf + TWIST*yf + BULGE*Rf Thus the distortions are equivalent to variations in crystal-to-film distance with position on the film. TILT is an xf-dependent variation due to rotation of the cassette about the film axis yf, while TWIST is yf-dependent. BULGE gives rise to a radial dependence of crystal-to-film distance and is equivalent to a cone-shaped film with a half angle close to 90 degrees. Note that whereas an error in crystal-to-film distance gives rise to a linear dependence of co-ordinate shift, distortions always result in a quadratic variation of discrepancy between observed and calculated positions on the film. For MAR image plate additional parameters TOFF and ROFF are refined by the integration program INTLAUE. @ REPLACE~THIS~PAGE~WITH~A~POSTSCRIPT~PRINT~OF~THE~FILE~coord1.ps @ REPLACE~THIS~PAGE~WITH~A~POSTSCRIPT~PRINT~OF~THE~FILE~coord2.ps