. INTRODUCTION~TO~THE~LAUE~SOFTWARE~SUITE~ =======================================~ . Daresbury Laboratory, October 1989 (+ later minor updates) @ 1)~INTRODUCTION~^ This document gives an overview of the Laue data processing software suite developed at the Daresbury Laboratory. A general overview of the development of the suite is given in reference 1. An account of the principles used, in the majority of the programs in the Laue software suite, is given in Helliwell et. al (reference 2). An application of the processing of Laue difference data is given in reference 3. It should be noted 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. 4,5). 2)~OVERVIEW~OF~THE~STAGES~IN~INTENSITY~MEASUREMENT^ This section and the next section outline the stages involved in the processing of Laue data and give the names of the main programs involved. The sequence of the processing stages, with an indication of the files required at the different stages, is shown in the flowchart in section 4. After scanning all the films in each pack, the stages in processing the data are as follows: a) NEWLAUE - Prediction of Laue patterns and determination of the crystal orientation. (Or more recently X-windows based program LAUEGEN) b) GENLAUE - Refinement of the cell and setting parameters for the crystal and making an accurate prediction of the reflection positions on the films. (Or more recently X-windows based program LAUEGEN) c) INTLAUE - Integration of the optical film densities at the predicted reflection positions for all the films in the pack. (Or more recently INTLDM) d) AFSCALE - Derivation of the inter-film scale factors for the pack, scaling together the films in a pack and applying Lorentz, polarisation and obliquity corrections. e) LAUENORM - Normalisation of the data to take account of the variation of incident intensity with wavelength and other wavelength dependent factors. This is usually done by using an internal scaling procedure based on symmetry equivalent reflections measured at different wavelengths. (An alternative method which involves scaling against a reference set of data may also be used where appropriate - LAUESCALE) f) UNSCRAM - 'Unscrambling' of reflection intensities which are harmonic overlaps (multiples). (Also DECONV and an option in LAUENORM) g) Scaling and merging of film packs - may be done as part of LAUENORM or externally to the suite using the CCP4 programs ROTAVATA/AGROVATA. 3)~PROCESSING~OF~DIFFERENCE~DATA^ When proccessing difference data, it would of course be possible to process each set of Laue data to a merged set of intensities (or F's) as described in the previous section and then scale the data sets together using one of the programs normally used in monochromatic work. However, an alternative strategy is available when Laue photographs can be obtained for the two crystal states with the crystal in the same orientation. Fractional differences are calculated and each reflection is individually scaled to the corresponding reflection in a reference set of data corresponding to the initial state of the crystal. This may be a monochromatic data set or a set of Laue data processed as described in the section above. The procedure is usually carried out by following stages (a) to (c) from section 3 followed by use of the program DIFFLAUE. @ 4) SUMMARY~OF~THE~MAIN~SEQUENCE~OF~PROGRAMS~USED~FOR~LAUE~DATA~PROCESSING^ . Start~ Notes^ | optional [.gen file] | [ ] indicates that file | was created by LAUEGEN~ NEWLAUE~ another program or Predict Laue patterns using the editor. Find crystal orientation | ( ) Optional output .gen file | files. | LAUEGEN~ GENLAUE~ .com/.ctl files contain Refine crystal orientation control data for the Predict all spot positions program in question. | .gen file | .gen and .vc files may .ge1/.ge2 files | be created using an if batch [.com or .ctl file] | editor if desired. | INTLAUE~ Integrate optical densities at predicted spot positions for all films in a pack | .ge1 file updated | | ---------------------------------------------- | | .ge1/.ge2 | .ge1 | [.vc file] | [another .ge1] | | [.ctl file] | AFSCALE~ UNSCRAM~ [.lcf file] | Scale and merge (+) Deconvolute | a film pack multiples DIFFLAUE~ | DECONV~ Calculate difference data .afout file | from two film packs with the (.vc file) | same crystal orientation and | scale to input data from the [more .afout]| .lcf file and get phases [.ctl file] | from that file | | LAUENORM~ .lcf file | Wavelength normalisation [more .lcf from DIFFLAUE] | of one or more pack | | DIFFLMRG~ .lcf file | Merge DIFFLAUE output files (shelx file) | | | .lcf file | Reflection data | eg h k l F sig(F) Reflection data for difference Fouriers: h k l F1 F2 phase . @ 5)~PREDICTION~OF~THE~LAUE~PATTERNS^ The positions of the reflections in a Laue pattern may be predicted given the cell dimensions, the crystal setting, the wavelength range lambda-min to lambda-max and the minimum observable plane spacing, dmin. In a prediction, it is assumed that lmin, lmax and dmin are sharp limits whereas in reality usually 'soft' limits. The program NEWLAUE is used for making such predictions. Its second main use is for solving the crystal orientation when the unit cell and crystal to film distance are known. The procedure involves matching the observed angles between the reciprocal lattice vectors for a selection of nodal reflections with a pre-calculated table of angles of potential nodal relections. The spot positions used may be obtained from a scanned film image using the program SPOTIN which is a cut down version of the program GENLAUE described in the next section. The orientation is defined in terms of missetting angles as used in the oscillation film processing package. The program is an interactive one and displays a menu of parameters and commands on the left hand side of the screen whilst reserving an area on the right hand side for the display of predicted patterns. Program options allow for pattern display, spot identification, pattern enlargement and hard copy prints of the generated Laue patterns and the program has also proved to be a valuable teaching tool. A more recent X-windows based program LAUEGEN now incorporates most of the features of NEWLAUE and SPOTIN (and GENLAUE see below). 6)~PARAMETER~REFINEMENT^ Before accurate intensity measurement can be attempted, it is essential to get a very close match between the predicted and observed reflection positions. The interactive program GENLAUE allows for reflection prediction, film display and parameter refinement. The parameters refined by the program are the position of the film centre, the crystal to film distance, the missetting angles and, optionally, the cell parameters. The user may cycle through the process of reflection prediction, film display and parameter refinement until a satisfactory fit is obtained. The observed reflection positions are obtained by displaying the digitised film image as a background corrected threshold plot. Reflections are identified using the cursor and their accurate positions are obtained from centre of gravity calculations. The refinement seeks to minimise the difference between the observed and calculated positions. The best reflections to use are the spatially isolated nodal reflections. Typically about 150 of these are used and a good fit should give an rms deviation of not more than 0.05 mm. A more recent X-windows based program LAUEGEN now incorporates most of the features of GENLAUE. 7)~INTENSITY~MEASUREMENT^ The problems of obtaining background corrected intensity measurements from Laue data are very similar to those experienced when using other camera geometries. For proteins, however, there may be additional problems as the reflections, especially along lunes of data, may be very closely spaced. The program INTLAUE based originally on the oscillation film processing program MOSFLM is used for measuring the intensities from Laue photographs. The program allows for the refinement of parameters, such as film twist, tilt and bulge, based on the the positions of well separated nodal reflections. Options are available for both box integration and for profile fitting. A profile fitting option, using learnt profiles, was introduced to improve intensity measurement in particular for the weaker reflections. The film is split into 5, 9 or 17 different regions for which profiles are derived from singlet reflections. Recently, much progress has been made by Greenhough et al. on the deconvolution of spatially overlapping spots. 8)~SCALING~WITHIN~FILM~PACKS^ Inter-film scaling factors of the following form have been assumed I1 = I2*exp(a*lambda**3) The program AFSCALE evaluates and applies the scaling factors. Three wavelength ranges are needed (lambda <= 0.49, 0.49 <= lambda <= 0.92, lambda > 0.92) because of the discontinuities in the scaling curve at the Silver and Bromine k edges, 0.49 and 0.92 A. The coefficient 'a' is evaluated for each range and refined using an iterative process. Some reflections predicted as singles (because their wavelength or d-spacing lies just outside the 'hard' limits assumed), in fact turn out to be multiples; these reflections can be detected at this stage as their I1/I2 ratios do not conform to those derived for the majority of the reflections, and are then rejected from the evaluation of the scaling coefficients. This affects particularly the scale factors at high lambda values. Once the scale factors have been determined, the data from the films are scaled together to give the intensities for the single reflections. The multiples are treated separately. Finally the data are corrected for Lorentz, polarisation and obliquity corrections. 9)~WAVELENGTH~NORMALISATION^ The program LAUENORM is used to derive a wavelength normalisation curve based on symmetry equivalent reflections measured at different wavelengths. The data are divided into wavelength bins and inter-bin scaling factors are derived from reflections which have symmetry equivalents in more than one bin. The bin scale factors, as a function of wavelength, are curve fitted and the relative scale factors for the individual reflections are derived from the curve. The scaling process is usually split into three separate wavelength ranges to allow for the film absorption discontinuities in the scaling function. If required, several packs of data may be handled simultaneously using an iterative procedure which alternates wavelength bin scaling and curve fitting with inter film pack scaling. In general, it is better to use crystals which are not as exactly set with an axis along the direction of the beam so that the off-diagonal terms in the overlaps matrix are more densely populated. For crystals with small unit cells there may be insufficient overlaps within a pack to derive a normalisation curve and in such a case it will be necessary to use several film packs measured for different settings of the crystal. If the crystal does not have symmetry equivalents then it should still be possible to use the method if multiple film packs for different crystal settings are used. 10)~UNSCRAMBLING~HARMONICS^ A complication of the Laue method is the presence of harmonics which are superimposed to form composite reflections on the film (multiples). The intensity of such a reflection is the the sum of the intensities of the constituent reflections. The integrated intensities of these multiples are evaluated as for the other reflections. The possibility of deconvoluting the harmonic components depends on the fact that different wavelengths are absorbed differently by the films and by any metal foils placed between the films. For a six film pack there is a theoretical limit of deconvoluting six superimposed reflections though in practice the limit is much lower as under and over exposed measurements must be rejected. Most commonly, only double reflections will be deconvoluted. The program UNSCRAM is used to carry out this unscrambling process and provide a set of data which may then be normalised as described above. A method for deconvoluting Laue harmonics data using direct methods in real space is also available (DECONV). Also there is an option built into the LAUENORM program which can deconvolute multiples data recorded on image using the varying nature of the wavelength normalisation curve. 11)~PROGRAMS~FOR~PROCESSING~DIFFERENCE~DATA^ The programs LAUEDIFF or DIFFLAUE are used to prepare difference data scaled on an individual reflection basis to a reference set of data. The coefficients are calculated as: . ( F(Laue_time2) - F(Laue_time1) ) Fdiff = --------------------------------- x F(Reference_time1) F(Laue_time1) . The reference e.g. native set of data could be a set of monochromatic data or a set of Laue data processed as described above and using the LAUENORM wavelength normalisation procedure. Phases in the reference data set may be passed on to the output set of data containing the difference coefficients. The preferred method of the two available is probably to use DIFFLAUE in which individual pairs of films, one from each pack, are scaled together and where the data from each of the pairs of films are merged as the final processing stage. In the alternative method, using LAUEDIFF (not shown on the flowchart), the film scaling within a pack is done using AFSCALE. LAUEDIFF then scales together the two packs and calculates the difference data as indicated above. It is important to remember that the crystal orientation must be very close for the two exposures so that the wavelength associated with each reflection is essentially the same (say within 0.1 Angstrom) for both measurements as the wavelength normalisation is effectively done on a reflection by reflection basis. If several film packs of data are processed using DIFFLAUE, the resultant data sets may be merged using the program DIFFLMRG. 12)~REFERENCES~^ 1) Campbell, J.W., Clifton, I.J., Elder, M., Machin, P.A., Zurek, S., Helliwell, J.R., Habash, J., Hajdu, J. and Harding, M.M. (1987) in Springer series in Biophysics, Vol 2., Biophysics and Synchrotron Radiation (ed Bianconi, A. and Congiu Castellano, A.),(Springer Verlag, Berlin, Heidelberg, New York, London, Tokyo) pp. 52-60 2) 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~ 483-497. 3) Hajdu, J., Machin, P.A., Campbell, J.W., Greenhough, T.J., Clifton, I.J., Zurek, S., Gover, S., Johnson, L.N. and Elder, M. (1987) Nature, Vol 329, 178-181 4) Arndt, U.W. & Wonacott, A.J., (1977). "The Rotation Method in Crystallography" Amsterdam: Netherlands. 5) Wonacott, A.J. (1980) Notes on a suite of programs for processing oscillation camera film data.