IUCr congress and general assembly

Structural phase transitions

The standing room only crowd attending the microsymposium on Structural Phase Transitions reflected the intense interest in this topic among crystallographers worldwide and the lively growth of activity on the subject during the last three years. In fact, many other microsymposia at the Congress included aspects of the field of structural phase transitions from different perspectives. This session focused on structural phase transitions and methodology.

The first lecture by C.J. Howard (Australia) highlighted the extreme efficiency of systematic use of group theoretical methods when analysing complex sequences of phase transitions in compound families. The use of tools such as those provided by the software package ISOTROPY (physics1.byu.edu/~stokesh//isotropy.html) by H.T. Stokes and D.M. Hatch (USA), was presented considering the particular case of some double and some cation deficient perovskites [Acta Cryst. B59, 463 (2003)]. Perovskites were also the subject of the second talk by D. Pandey (India), more specifically, the morphotropic tetragonal-rhombohedral boundary in the phase diagram of relaxor ferroelectrics. The existence of a difficult to characterize intermediate monoclinic phase in PZT has been the subject of intense discussion and has become a central feature for understanding the high piezolectric response of these sytems at this particular composition. M. Braden (Germany) presented a review of the properties of the system [Ca,Sr]RuO₄ as a function of composition and temperature. In spite of the isovalent character of the substitution, the system exhibits a rich variety of physical phenomena: unconventional superconductivity, a Mott-insulating phase, one of the highest Cp/T low-temperature ratios observed in a transition metal oxide and a metamagnetic transition. The smaller ionic radius drives a series of structural phase transitions which are intimately coupled with physical properties. As a function of Ca⁺² concentration the octahedra exhibit tilts and rotations which strongly reduce the electronic hybridization thereby decreasing the electronic band widths. In this sense the increasing structural distortions enhance the role of electronic correlations. Strongly distorted Ca₂RuO₄ is even an antiferromagnetic Mott-insulator. Although centered on neutron diffraction and inelastic scattering, the talk reviewed the results of a remarkable variety of experimental methods, showing the power of combining different techniques when interpreting such complex systems.

T. Chatterji (France) reported an isosymmetric transition in LaMnO3 with a discontinuous volume reduction, which was compared with the ice melting transition [Phys. Rev. B 69, 132417 (2004)]. The final talk by S. Leoni (Italy) covered molecular dynamics studies of reconstructive phase transitions. By means of a novel method which samples transition paths, S. Leoni et al. have been able to explore the most probable transition routes for this type of transformation. This work introduces the missing link between homogeneous transition paths studied theoretically and the local nucleation and growth mechanism expected in a real system. He presented the wurtzite to rocksalt transition in CdSe as an example [Phys. Rev. B72 064110 (2005)]. Inhomogeneous transient states exhibit local atomic configurations, as interfaces between the two end states, which can be identified with some of the transformation paths proposed, based on theoretical considerations. Thus for the pressure-driven transition of ClNa from the B1 (ClNa) type to the B2(ClCs) type structure, the authors could observe a local B33 (α-TlI) configuration interface between the two end structures [Phys. Rev. Lett. 92, 250201 (2004)], in accordance with one of the transition paths proposed for this transition.

J. Manuel Perez Mato and Ulrich Bismayer

Instrumentation for next generation X-ray sources

The topic of the microsymposium was the next generation of X-ray sources, in particular, X-ray free-electron lasers (FEL). Three X-ray FEL projects for hard X-ray radiation down to 0.1 nm wavelength using the self-amplified spontaneous emission (SASE) scheme are currently in construction, the Linac Coherent Light Source (LCLS) at Stanford, USA, the Spring-8 Compact SASE Source (SCSS) at Harima, Japan, and the European XFEL at Hamburg, Germany. The properties of X-ray FEL radiation are unique in that 10¹² - 10¹⁴ photons are provided by a single pulse of typically 100 fs duration, in a bandwidth of about 0.1% of the photon energy. The radiation is furthermore transversely coherent. These properties translate into a peak brilliance about nine orders of magnitude higher than the best synchrotron radiation sources today. New and currently-impossible scientific applications have been proposed for these facilities, which will impose tough requirements on the experimental equipment. X-ray FEL experiments are expected to differ substantially from experiments at synchrotron sources. The five speakers of the session presented the state-of-the-art for different aspects of instrumentation for X-ray FEL experiments.

K. Gaffney (USA) discribed investigations of ultrafast processes on the sub-picosecond timescale using pump-probe experiments. In this type of experiment a pump pulse excites the sample and the X-ray pulse is typically used as a probe to observe the resulting rapid structural changes. The pump pulse could be either optical or X-ray, leading to different requirements for instrumentation. The use of optical lasers requires establishing a time correlation between the laser and X-ray pulses at the femtosecond level. This could be achieved either by synchronisation or by measuring the delay for each pair of pulses and appropriately tagging the data. The latter approach has been successfully tested at the Sub-Picosecond Pulse Source at SLAC, yielding a time resolution of the order 150 fs for laser-pump/X-ray-probe experiments. The imaging of single molecules using intense, ultrafast X-ray pulses was presented as a second X-ray FEL experiment by S. Marchesini (USA). A large number of new, difficult issues will need to be resolved: sample injection, preparation of very short X-ray pulses, absorption of X-rays in the molecules and subsequent radiation damage, and the classification of data sets and reconstruction of the 3D structures. Recent experiments using soft X-ray radiation achieved 3D reconstruction of cells and test objects at high resolution and showed that the algorithms are available.

M. Yabashi (Japan) described the role of optical elements in preserving the high degree of coherence of XFEL radiation. State-of-the-art techniques suggest that optics for coherent X-ray radiation at 0.1 nm wavelength are readily available.

B. Lengeler (Germany) presented the latest achievements using refractive lenses and discussed the limitations for observing very small focii.

Finally, C. Brönnimann (Switzerland) presented the advances in the field of pixel detectors for synchrotron radiation applications and discussed necessary modifications and possibilities for XFEL experiments with far higher pulse intensities.

Th. Tschentscher and J. Arthur

Advances in computational methods for powder diffraction microsymposium

This microsymposium highlighted computer software advances in powder crystallography. Andrew Wills (UK) described a generalized approach to magnetic symmetry, using irreducible corepresentations that allows both commensurate and incommensurate magnetic structures to be modeled. He has implemented this complex theory with a friendly interface in a program called SARAh, which is freely distributed. Juan Rodriguez-Carvajal (France) presented his work on a huge open source subroutine library for crystallographic calculations, CrysFML, which is written in object-oriented Fortran (F95). The FullProf Rietveld program already uses parts of CrysFML and it is progressively being reimplemented to take increased advantage of the modular structure of this library. Simon Billinge (USA) discussed data analysis & modeling for total scattering and goals for a new generation of diffraction analysis envisioned as part of the DANSE project, which has been proposed to the National Science Foundation. Gavin Vaughan (France) showed how advanced computational analysis and superb microbeam diffraction capabilities allows indexing of individual crystallites in a powder diffraction sample. This allows single crystal diffraction intensities to be measured directly from powders. Finally, Samantha Chong (UK) presented her thesis work (under advisor Maryjane Tremayne) on structure solution strategies using powder diffraction.

Brian H. Toby

The big question in structural genomics: discovering function from structure

Given that more than 50% of the proteins encoded by any genome are of unknown or uncertain function - to what extent can the protein structures contribute to the discovery of function? The speakers in the microsymposium, which was chaired by Ted Baker (New Zealand) and Sung-Hou Kim (USA), reflected the global spread of interest in this topic: Bill Shepard (France), Osnat Herzberg (USA), Ray Stevens (USA), Seiki Kuramitsu (Japan) and Al Edwards (Canada, UK). Talks ranged from detailed analyses of a single protein to discussions of broad biochemical and biophysical screening assays.

Bill Shepard described a hypothetical protein from Mycobacterium tuberculosis that proved to be a fatty acid binding protein, although its true physiological substrate has yet to be fully identified. Osnat Herzberg showed that time and expert knowledge can be important allies. One hypothetical protein from Haemophilus influenzae was almost a 'singleton' when the structure was first solved, but is now representative of a family of more than 800 members, with both structure and function (as a tRNA editing domain) established. She also showed the importance of structure in differentiating enzyme functional families through specific structural signatures, even in the presence of high sequence identity.
Ray Stevens showed that not even nature's thunderbolts can keep a good talk down. When Jerry Joseph, the invited speaker, was delayed by thunderstorms in Georgia, Ray Stevens stepped in at 30 minutes notice, and gave a fascinating account of their SARS structural genomics efforts. It is remarkable how the technological advances of recent years now allow structural biology to address new viral threats with great urgency - although Ray also cautioned that the biology still remains the biggest challenge in such cases, especially where protein-protein interactions need to be elucidated. Seiki Kuramitsu described progress with the RIKEN structural genomics project on an extreme thermophile, which has the ambitious long-term goal of in vitro reconstitution of the organism, as well as characterisation of its proteins.
Finally Al Edwards finished the microsymposium in rousing style, with a talk that included a surprising excursion into the mysteries of soccer (revealing to this listener some unexpected physical similarities with ice hockey). The human structural genomics project he heads, which now incorporates sites in Toronto, Oxford and Stockholm, and which operates as a public good enterprise, is currently determining 13 new protein structures per month. Importantly, it has developed a range of screening assays, for example to discover enzymatic activity or ligands to assist in crystallization, which are clearly a very important resource for the community.

Ted Baker

Putting the pedal to the metal: speeding up the structure determination process

One major task of structural genomics is the design of tools that increase the efficiency and speed of experimental structure determination. Such developments will ultimately result in a reduction of the average cost per structure for the more standard protein targets while allowing efficient screening of many hundreds of crystals. The implementation of high-throughput protein expression and crystallization robots has already done much to accelerate the wet-chemistry part of a structure determination.
MS23 focused on ongoing efforts to automate the X-ray data collection and phasing parts of the structure determination process. Martin Walsh (France) gave an overview about the hardware and software implemented at the BM14 beamline (ESRF, Grenoble) and the e-HTPX project that will enable users to manage the entire data collection process with a seamless exchange of wet-chemistry and diffraction data over the internet. Wayne Anderson (USA) reported on the Automated Crystallography System (ACrS). The system, which was successfully applied in 75% of all cases, relies on existing software and explores different software combinations in parallel. The application of ACrS was discussed in light of two crystal structures that were solved using the anomalous signal of zinc, since Se-Met MAD phasing was impossible. John Rose (USA) described a 24-hour synchrotron shift during which five structures were determined by SAD using the SCA2 structure pipeline. This pipeline operates either in an automatic hands-free mode or at an expert user level. In contrast to this the HKL2000_ph program combines data reduction and phasing into a single stand-alone program. Wladek Minor (USA) illustrated the capabilities of this program by solving SAD and MAD structures on stage. Keith Wilson (UK) focused on ongoing efforts to streamline the entire structure determination process from the perspective of medium sized structural biology laboratories. Lessons were drawn from a recent workshop on Structural Proteomics in Europe with particular examples of high-throughput structure determinations of Bacillus anthracis proteins. These systems are already accelerating X-ray structure determination in an impressive way and facilitate structure determination even by non-expert users.

Peer Mittl and Yvonne Jones 

Structure determination from powder diffraction data (organics)

The prevailing methods for structure solution of organic materials from powder diffraction data exploit known molecular structure via a global minimisation in real space, though direct methods remain an essential option.

Vincent Favre-Nicolin (France) gave an expert overview of his freely available FOX program. Originally developed for inorganics, its versatility allows molecular structures to be solved via Monte Carlo or parallel-tempering minimisation routines. Molecules are described and manipulated using bond-length, bond-angle and dihedral-angle restraints, providing a very flexible way to define the degrees of freedom that need to be optimised. New features include maximum likelihood which enlarges the hypersurface near the global minimum and is useful when the model is incomplete or has systematic errors. Vincent showed some impressive examples, including two triglycerides from chocolate with up to 57 torsion angles that took two months of computing time to solve.

Many structures are solved from laboratory powder data, and the question of their accuracy can arise, especially if the structure does not agree with the expectations of the synthetic chemist. Vladimir Chernychev (Russia) related such an instance when the organic ligand was clearly tridendate, whereas the chemist believed that it should be bidentate. The controversy was only resolved once a single crystal was finally grown, confirming the powder structure. Further examples were given, corroborated by synchrotron or neutron diffraction data, to demonstrate the veracity of structures determined from laboratory data.

Cikui Liang (USA) explained how direct-space methods can fail if the powder pattern is of inadequate quality, or if there are too many degrees of freedom to be defined. The structure must make chemical sense, so inclusion of potential energy in the minimisation can help, e.g. with a close-contact penalty function and a combined figure of merit. Energy can also be incorporated into the Rietveld refinement provided accurate potentials are known. The relative weight to be given to the diffraction data versus the energy is not intuitive, but can be derived via a Pareto optimisation to find the proper balance between fitting the diffraction pattern and minimising the energy. These ideas were illustrated with some nice examples, e.g. (E)-2-(4,6-Difluoroindan-1-ylidene)acetamide, analysed using Materials Studio.

Philippe Fernandes (UK) reported attempts to use direct methods and differential thermal expansion (DTE) to solve the structure of γ-carbamazapene, with 72 non-hydrogen atoms. With powders, data to atomic resolution is difficult to obtain, especially with severe peak overlap, but harnessing DTE can improve things, because peaks overlapped at one temperature may be better resolved at another. Data collection using synchrotron radiation was concentrated at high angles to improve the quality of this crucial part of the pattern. To minimize radiation damage, low temperatures and a wavelength of 0.8 Å were used, automatically translating and changing capillaries to expose fresh sample to the beam. All in all an impressive amount of data, from which a set of intensities was extracted using a multipattern Pawley procedure implemented in Topas. Direct methods have not yet been successful, but work is continuing. Using DASH and global optimization, the structure could be solved and refined without additional restraints. Anders Markvardsen (UK), standing in for Kenneth Shankland, emphasized that computing power is essential for solving complex structures by multisolution simulated annealing (SA). Whereas, small molecules may need only seconds per SA run, and have close to 100% success rates, complex structures need possibly hours per run, and success may fall to below 5%. Implementing a grid system can give access to almost unlimited computing power, by using unused time on other computers. Anders illustrated the impressive performance of this multi-machine approach, combined with the hybrid Monte Carlo search algorithm. The grid does not just give you more, faster: it allows you to attempt the otherwise impracticable.

Andy Fitch and John Faber

Recent advances in quasicrystal research

Rónán McGrath (UK) reviewed recent experimental studies of surfaces exhibiting fantastic long-range order on the atomic scale, as can be seen in the beautiful STM image below. He also showed that a vibrant current research theme is the use of quasicrystal surfaces as templates for the growth of novel nanostructured systems. Zorka Papadopolos (Germany) showed how accurate STM data can be used to determine the principles that govern with which surfaces quasicrystals terminate.
A common theme of the talks was the increasing quality and accuracy of quantitative information obtained in current experiments. Not only is it possible to reach atomic resolution in real-space images, it is also possible, as shown by Marc de Boissieu (France), to obtain extremely precise data from X-ray diffuse scattering in order to extract quantitative information regarding the role played by phasons in the thermodynamics of quasicrystals. The high precision of current X-ray diffraction experiments allows the study of quasicrystals at extreme pressures and temperatures, as described by Günter Krauss (Switzerland). To the astonishment of all participants, Krauss reported that a number of different decagonal quasicrystals remain stable, without any structural transformation, up to tens of GPa and almost 1000 K. These data are sure to keep theorists busy for some time.
The microsymposium ended with a presentation by Yushu Matsushita (Japan), who introduced the newest sibling in the family of quasicrystals - a polymeric quasicrystal. The building block of this system is composed of a star-shaped configuration of three polymers. By varying the length of one of the three arms, the system exhibits a variety of long-range ordered 2-dimensional structures with tiles on the order of 100nm. Among these is an approximant of the dodecagonal square-triangle tiling, which with some tweaking may yield an exact dodecagonal quasicrystal in the near future. This possible new addition to the growing number of systems exhibiting quasiperiodic long-range order is yet another indication that quasicrystals will remain in the forefront of crystallographic research for years to come.

Ron Lifshitz and Koh Saitoh

Detectors: developments and requirements for X-ray, synchrotron and neutron sources

The first speaker, Burckhard Gebauer (Germany) described several different types of neutron detectors developed in DENTI, using MSGC and Si-MSD. Considerable efforts are being made for ASICS development. Three different types of X-ray area detectors were described. Jules Hendrix (Germany) talked about a Se-based TFT flat panel detector with a diagonal size of 555 mm. Data obtained using synchrotron radiation were presented. He stressed the high spatial resolution due to the use of direct conversion by Se. The detector was to be commercially available for protein crystallography by the end of 2005. Gregor Huelsen (Switzerland) presented the recent status of the PILATUS detector. A detector developed in colloboration with a synchrotron facility, it is a silicon pixel array detector using bump-bonding and modern semiconductor technology. A tiled assembly of the 1M prototype has been tested for protein crystallography. After corrections for inhomogeneity and distortion, an electron density map was successfully obtained. Roger Durst (USA) announced Resistive Microgap Detector, which is a gas detector with a high local counting rate. The readout is through a delay line. For high gain and stability, electrodes were specially designed to avoid sparks. Parallax problem was solved by field shaping. The latter two X-ray detectors are photon counting, and David Laundy (UK) described his analysis of dead-time when such detectors are used with a pulsed X-ray source like synchrotron radiation. This problem will be more serious in future because the proposed next generation x-ray sources are all pulsed sources.

Naoto Yagi and Christian Broennimann

Structure determination from powder diffraction data (inorganics)

Powder diffraction was used primarily for phase identification until the mid nineteen nineties. Recent rapid progress in X-ray sources, instrumentation, computers, software and methodology has made the use of powder diffraction for solving crystal structures realistic.

Textured powder samples can be exploited to resolve the problem of overlapping reflections in powder diffraction. Lars Kocher (Switzerland) showed an improved experimental set-up collecting only a quarter of the full 360° scattering pattern on an image plate system gathering data to higher 2θ values. He presented algorithms which help reduce the disadvantage of poor statistics when using a new linear detector with extremely high resolution.
Anna Grazia Moliterni (Italy) provided an overview of the 'EXPO2004' software package and described new features of EXPO2005 which includes improved methods and algorithms for space group determination, background calculation, estimation of integrated intensities and phasing. Moreover, a new global optimization approach exploiting Fourier maps provided by direct methods has been optimized for organic structures.
Robert Dinnebier (Germany) demonstrated that many phases may still be missing in the phase diagrams. He presented powder diffraction studies at variable pressure and temperature conditions. Using a new fast image plate detector he was able to dramatically improve data collection, leading to the solution of the crystal structures for many intermediate phases. Elaborate strategies were adopted for processing large amounts of raw area detector data to obtain the final individual diffraction patterns.

Thierry Bataille (France) talked about the ab-initio solution of crystal structures of products obtained by thermal decomposition. Due to phase transitions during heating/cooling of the samples and the corresponding reorganization of the crystal structures, single crystal data are generally not available for these compounds. He described in situ powder data collection using Bragg-Brentano optics in combination with a conventional X-ray source and presented several new crystal structures.

A special highlight was Pavol Juhas' (USA) description of an elaborate new algorithm to reconstruct the three-dimensional structure of a single-element molecule from its pair distribution function (PDF). Step by step the build-up of clusters starting from individual atoms was demonstrated. New atoms were added to the cluster, controlled by a figure of merit based on the agreement between the calculated and observed PDF. Several trials were ranked according to the figure of merit and a special new algorithm inspired by soccer leagues was applied. The case of a C60 molecule was analyzed.

Angela Altomare and Holger Putz

Charge, spin and momentum densities in materials science

This session presented a wide range of scientific problems that can be favorably addressed with 'sub-atomic' crystallography through, for example, very accurate low-order structure-factor measurement. Experimental techniques spanned high energy and high intensity synchrotron X-ray diffraction, through convergent-beam electron diffraction (CBED) to the measurement of magnetic Compton profiles. Thomas Lippman (Germany) opened the session by discussing the virtues of very high energy radiation (> 100 keV) to minimize extinction errors for obtaining charge densities (CDs) on inorganic structures, and he exemplified the vast potential of the technique with a study of charge and orbital ordering in heavy atom oxides. Part of the talk was devoted to describing the new GKSS high energy beam line being constructed at HASYLAB and the great promises of the new PETRAIII ring. Benoît Guillot (France) discussed an amazing study of the 6700 atom human aldose reductase protein, giving a detailed description of how to handle millions of reflections in the multipole model - and it actually worked. Convincing electron deformation maps can now be retrieved even on the inhibitor, and subtle electronic features never before exposed in such large molecules can be analyzed. In an eminently scholarly talk Dietmar Stalke (Germany) then took the audience back to freshman chemistry and challenged the concept of hypervalency. Applying the quantum theory of atoms in molecules on the experimental CDs of a range of formally hypervalent compounds our foundations were shattered and then rebuilt in simple bonding models, which have also led to the development of new types of chemical reactions. Very predictive indeed! Jian-Min Zuo (USA) addressed the problem of data accuracy in vital low order reflections, which can be overcome by CBED using a nanometer probe and perfect crystal theory to avoid extinction errors. A range of 'simple' inorganic materials were discussed, and a detailed comparison with state of the art theoretical calculations showed that not far beyond monoatomic crystals, theory gets seriously challenged. Experimental data do in fact have smaller errors than the differences between theoretical models, and CD studies are important for further development of ab initio methods. In the concluding talk Yoshiharu Sakurai (Japan) explained about the peculiar material UGe₂, which has coexistence of superconductivity and ferromagnetism. The magnetism in the system was approached by a novel orbital decomposition scheme of the magnetic Compton profiles. The method allows the temperature dependence of the separate orbital populations to be studied, and this led to the conclusion that superconductivity in UGe₂ is mediated by orbital fluctuations. In summary, the session spanned from simple metals to proteins, and from small organics to ultra heavy inorganics on the journey through charge, spin and momentum space. Development of new experimental techniques continues to propel the vibrant field of sub-atomic crystallography, which is clearly a very.

B. Iversen and J. Spence

Enzymes and allostery

This microsymposium was organized through a joint effort of Alexander Wlodawer (USA) and Silvia Onesti (UK). Unfortunately, the latter was not able to attend due to her involvement in an applied genetics experiment, the results of which were due in late October, but both co-chairs were responsible for the difficult decisions involved in choosing only five speakers from among many excellent abstracts. The session was started by Lukasz Lebioda (USA) who presented exciting results on the negative cooperativity of human thymidylate synthase. That enzyme cycles between an active and inactive conformation, both of which may be independently targeted by inhibitors. The inhibitors that stabilize the inactive conformation show positive cooperativity with currently used antifolate cancer drugs. Momoyo Ishikawa (Japan) discussed substrate channeling in a fatty acid β-oxidation multienzyme complex. Large changes in the structure of the enzyme are responsible for sequential presentation of the substrates for several stages of catalysis, without their intermediate release. Pamela Williams (UK), a finalist in this year's Pharma Achievement Award, gave a fascinating talk on structural studies of cytochromes P450, enzymes directly responsible for metabolism of a majority of drugs. Cytochromes such as CYP3A4 and CYP2C9 have been studied as complexes with a variety of marketed drugs and their structures provide novel insights into the principles of substrate binding and allostery. Wojciech Rypniewski (Poland) presented the first structure of a eukaryotic phosphofructokinase 1, describing how catalytic sites could have evolved into allosteric effector sites. The final speaker, Marcus Rudolph (Germany) apologized that his favorite macromolecule, formylglycine generating enzyme, is not actually allosteric, nevertheless the combination of high resolution crystal structure and extensive mutagenesis went a long way towards explaining the molecular basis of multiple sulfatase deficiency, an inherited human disease. Thus the session ended up being 100% enzymes and 80% allostery, a combination that seemed to be of considerable interest to the audience.

Alexander Wlodawer

Applications of synchrotron and neutron facilities in structural chemistry

In this microsymposium some recent structural highlights obtained by chemical, biological and crystallographic groups worldwide, using both synchrotron and neutron diffraction facilities were reported.

Russell Morris (UK) described fascinating structural studies using a combination of microcrystal X-ray diffraction techniques, carried out on Station 9.8, at the SRS, UK, coupled with laboratory based MAS NMR studies of zeolite structures. Some of these zeolite materials exhibit negative thermal expansion properties, and through structural analyses it has been possible to explain these temperature effects by librational changes in the cage structures.Wolfgang Scherer (Germany) showed how single crystal neutron diffraction studies, augmented by high resolution X-ray data, could be used to provide an accurate picture of the C-H and M-H interactions in a series of catalytically important metal complexes. The experimental pictures did not always correspond to preconceived predictions and the results shed new light on 'agnostic' interactions in a range of main group and transition metal complexes.
Alberto Albinati (Italy) showed how single crystal neutron diffraction studies, carried out at the ILL, Grenoble, could be coupled with inelastic neutron scattering studies to provide new information on the bonding in classical and non-classical metal hydride complexes. This detailed information could not be obtained by other experimental techniques and contrasted with solid state 1H NMR results. Alberto emphasized the importance of carrying out librational corrections in this type of analysis.
Bob Bau (USA) presented a 'hot off the press' result, the structure of the first complex with a four-coordinate interstitial hydride ligand. This has been one of the 'holy grails' of cluster chemistry for three decades, and in a single crystal neutron diffraction study of Y₄H₈(Cp'₄)(THF) [Cp' = C₅Me₄(SiMe₃)] one of the hydrides sits in the middle of the metal tetrahedron.
Matthew Blakely (France) emphasized the importance of neutron diffraction in locating hydrogen/deuterium atoms in protein structures and described a method that he and his co-workers had developed for cryo-cooling crystals of a suitable size for neutron diffraction studies to 15K. Structure analysis using data at this low temperature gave much improved structural refinements and provided more information as to the nature of the hydrogen bonding in several systems compared to room temperature data.

Paul Raithby and Trevor Forsyth

Microstructural properties from powder diffraction data

The first lecture 'Subgrain Size Distributions, Dislocation Structures, Stacking- and Twin Faults and Vacancy Concentrations in Crystalline Materials Determined by X-ray Line Profile Analysis' by Tamás Ungár (Hungary) presented many different applications of line broadening analysis based on physically sound models. Especially interesting were new insights into the relationship between measurements of domain/crystallite size, as obtained by X-ray diffraction and transmission electron microscopy, and analysis of vacancy concentrations. Matteo Leoni (Italy) gave an overview of different types of defects and modeling of their influence on diffraction lines. A theoretical overview was followed by several examples treated by the program for modeling line broadening, WPPM. Radomír Kuzel (Czech Republic) discussed defect structure obtained by high-pressure torsion and films made of colloidal gold nanoparticles. Line-broadening analysis was complemented by diffuse scattering studies, TEM, and life-time positron annihilation spectroscopy. Nicolae Popa (Russia and Romania) discussed 'Size Anisotropy and Lognormal Size Distribution in the Powder Diffraction Whole Pattern Fitting'. He proposed a new model that uses ellipsoids to model size anisotropy of crystallites, compared it with the model that employs spherical harmonics, and tested it on a zinc-oxide powder. The last speaker, Nicholas Armstrong (Australia and USA) described the 'Development of a NIST SRM 1979 Nano-Crystallite Size Standard for Broadening of X-ray Line Profiles'. Especially interesting was the certification process of the standard reference material (SRM), which employed the Bayesian/maximum entropy method.

Davor Balzar and Paolo Scardi

Structures, phase transitions and properties at high pressure

Structures, phase transitions and properties at high pressure have been the center of high-pressure research for decades, since they provide fundamental structural information that is the basis for all high-pressure science, from biology to cosmology.

In recent years, enormous progress has been achieved in experimental techniques, in particular, in X-ray and neutron diffraction under extreme pressure-temperature conditions. A remarkable example of high pressure single crystal work was presented by Tetsu Watanuki (Japan) on subtle structural modification of a complex Cd-Yb 1/1 quasicrystal approximant. This was clearly seen in his synchrotron X-ray oscillation photographs showing the evolution of superlattice reflections with decreasing temperature at high pressure. This structural change is attributed to the successive orientational ordering of the Cd₄ tetrahedron which is the most inner unit of the four-shell atomic clusters occupying the bcc lattice sites. These four newly found structures, together with two known structures, constitute a complex P-T phase diagram.

Nozomu Hmaya

Advances in computational methods for protein crystallography

This session covered developments in methods, which will be useful in the investigation of more difficult problems not amenable to automated solution and in future high-throughput work.

Randy Read (UK) presented 'New likelihood based phasing methods in 'phaser''. He introduced likelihood phasing, in which the best model is the one which is most probable to produce the observed data. This approach has been applied to both molecular replacement and heavy-atom phasing in 'phaser' software.

For molecular replacement, this software provides a number of benefits: likelihood-enhanced rotation and translation function provide greater sensitivity; anisotropic data correction helps with some difficult datasets; and modular programming allows the implementation of complex automation strategies which involve the testing of many chains of hypotheses when searching for multiple models. An example was given in which a 4-helix bundle was solved in under a minute by exploring permutations of helical fragments.

Heavy atom phasing for SAD and MAD experiments has also been implemented, using the conditional probability of the Friedel mate to modify the phase probability distribution for the original reflection. Log-likelihood gradient maps from this approach are very powerful in identifying minor sites, for example in the case of Halide soaks. A high resolution case was presented in which this method allowed the location of Fe, S, and eventually the C atoms within the phasing program.

In the future, it is hoped that molecular replacement and heavy-atom phasing will be combined for difficult problems. Problems still remain in the exploitation of non-crystallographic symmetry in molecular replacement. Asked about the limits of the method, Read said models may be as small as 15% of the total structure, or as low as 20% sequence homology.

Marc Schiltz (Switzerland) presented a lecture on 'Broken Symmetries in Macromolecular Crystallography', dealing with cases when multiple measurements of symmetry related reflections give different results.

The first case considered was that of radiation damage, which over the course of the experiment causes changes in the crystal leading to changes in the measured intensities. Others have employed this as a means to determine a model of the changes within the crystal. The focus of this work was the use of this information to enhance phasing. Two parameters are added to each heavy-atom to model change in occupancy during the experiment.

A more unusual application of this approach arises when polarized X-rays are anomalously scattered from heavy atoms with anisotropic density distributions, for example due to bonding effects. In this case, the magnitude of the anomalous effect may be dependent on the incident and diffracted beam direction. Schiltz was asked about the effect of detector size on the diffracted beam direction, and he suggested that the incident beam is more important in this effect.

Hongliang Xu (USA) presented his work on 'Statistical direct methods of phase determination'. Direct methods may be approached by optimizing some target function whose value is minimised by the true phases. The assumption of a uniform random distribution of atoms leads to the result that triplets of structure factors whose normalised magnitudes are large are likely to have phases which sum to zero. This principle has been employed in the tangent formula, and in the minimal function used in the Shake-and-Bake direct methods software. However the effect becomes weaker as the number of atoms increases.

A new approach has been implemented in which phases are shifted to increase the fraction of the triplets whose phases fall within some range from zero. Analysis of data from real problems suggests that the width of this interval should increase as the number of atoms in the problem increases. This approach improves the success rate in determining heavy atom substructures over a range of test cases. It was suggested that the adoption of a linear integer method increases the radius of convergence for the calculation.

Pavol Skubak (The Netherlands) presented his work on 'Direct use of SAD phase information in automated model building and refinement', implementing model refinement directly against anomalous data in the REFMAC5 software, comparing three approaches. Traditionally, only the estimated non-anomalous magnitudes were used in the refinement. More recently, phase probabilities have been output from the phasing software, in the form of Hendrickson-Lattman coefficients, and employed in refinement software to provide more information from which to determine the model parameters. There are problems with this approach, in that the heavy atom phase distribution is not updated to include information from the final model, and the dependence between the heavy atom and model phases is ignored. To address these problems, all of the SAD data is used in the refinement program. The software was tested on a range of problems and was as good as existing methods in every case. It provided a significant benefit, measured by residues built in ARP/wARP, in about 30% of cases.

R. Read and G. Bricogne (UK) suggested that the anomalous phasing information was well described by Hendrickson Lattman coefficients, so the improvements may arise from elsewhere in the process.

Ralph Grosse-Kunstleve (USA) presented a lecture titled 'Hybrid Programming in Crystallography: Phenix.refine and phenix.hyss'. The 'phenix.refine' macromolecular refinement software uses components developed in the PHENIX project including bulk solvent model, and restraints on the isotropic B-values. A grid search of bulk solvent parameters avoids some problems of instability in the calculation and restraint of B-factors by proximity is found to be helpful in avoiding extreme values.

The 'phenix.hyss' software is a substructure solution program combining Patterson and direct methods with dual-space recycling. The software is open-source.

Piet Gross and Kevin Cowtan 

Chemical insights from electron density studies and wavefunctions

This microsymposium ranged from fundamental to applied aspects of the electron density in molecules and crystals. In his keynote lecture, M. Spackman (Australia) showed how the number of publications of charge density studies increases year after year and he enumerated the accurate electrostatic properties that can be derived from high resolution X-ray diffraction experiments for molecules and solid materials. R. Hoffmann (USA) discussed theoretical approaches for the characterization of the molecular bonds by using Bader's topological theory and D. Jayatilaka (Australia) showed how wavefunctions can be fitted to experimental diffraction data. The first speaker in this microsymposium, Richard Bader (Canada), offered deep insight into the nature of chemical bonding and its concept in chemistry. He redefined the basis of quantum mechanics in relation to the properties of atoms in a chemical system through the topological properties of its electron density. He demonstrated how the resulting boundary conditions can be useful for the extension of his theory to open systems. Jean-Michel Gillet (France) showed the possible combination of theory and experiment in a joint refinement to reconstruct reduced density matrices for small to extended molecular systems. Peter Luger (Germany) discussed the use of synchrotron radiation to collect very accurate data suitable for electron density studies. The application to strained carbon ring and cage systems with conformations close to that of fullerenes was highlighted in this presentation. He showed how topological analyses of the electron density for such compounds are powerful tools to characterize even unusual carbon bonds. So-called hypervalent compounds containing Si and Ge atoms were presented by Alexander Korlyokov (Russia). Examples of Si(Ge)...O(N) hypervalent bonds were discussed in the scope of calculated and experimental electron density topological properties at the bond critical points. Anders Madsen (Denmark) showed how accurate electron density determination and rigid-body TLS analyses can be pertinent to reveal physical properties of materials. On this basis, he could explain the observed difference in calorimetric and energetic results of diastereomeric xylicol and ribitol compounds. He illustrated his talk with animated videos showing the harmonic and uncoupled thermal motion of each molecule in the crystal lattice.

N.E. Ghermani

Protein interactions with other biological macromolecules

E. Yvonne Jones (UK) discussed strategies to produce X-ray-quality crystals of components of the extracellular region of the receptor protein tyrosine phosphatase mu. These strategies have recently led to the structure of the full-length protein including four fibronectin domains which are seen to adopt an extended 'beads-on-a-string' conformation.Chris Hill (USA) discussed so-called molecules of mass destruction, i.e. protein-degrading complexes. He focused on various activators of proteosomes and how they regulate access to proteolytic sites within the proteosome interior.
The RecBCD protein is a multifunctional enzyme complex that provides one mechanism for the repair of doublestranded breaks in DNA. Dale Wigley (UK) discussed the large and complicated, yet modular interdependent, structure of the enzyme. The RecBCD structure explains the regulation of nuclease digestion and also suggests how the protein can recognize single-strand DNA sequences as they pass through the complex.
Yong-Seok Heo (Korea) discussed the structures of different scaffolding proteins (JNK1, 2 and 3) that assemble components of the c-jun N-terminal kinase signaling pathway. Complexes of human JNK1, JNK2, and JNK3, with their appropriate activating peptide, provide a structural rationale for the specificity of peptide recognition and how associated hinge motion allosterically modulates ATP binding.
Hanna Yuan (Taiwan) described a number of structures bearing the so-called H-N-H motif that forms the core of the active site of many endonucleases involved in restriction, repair and degradation of DNA. One of these proteins is the bacterial toxin ColE7. The structures of the nuclease domain of this protein in complexes with various DNA duplexes were described. The results show how the H-N-H motif is bound in the minor groove and primarily contacts phosphate groups rather than ribose groups or bases. The His-metal fold of the active site suggests a universal mode of DNA recognition and interaction.

B.W. Matthews and Z. Rao

Topology of crystal structures: nets, knots and surfaces

Interest in coordination polymers, metal-organic frameworks, and supramolecular networks is rapidly expanding not only for their potential properties as functional solid materials but also for their intriguing architectures and topologies. Indeed, many examples have evidenced structural types and topological features unprecedented in the world of inorganic compounds and minerals, but a unified view of the observed complex topological phenomena is still lacking. Mike O'Keeffe (USA) illustrated the natural tiling approach to describe nets and that the self-dual tilings show the full catenation of the most frequently interpenetrated nets and the relation to the five genus-3 minimal surfaces was shown. Stephen Hyde (Australia) explored in depth the relation between the minimal surface and the hyperbolic plane showing results from the 'Epinet' project (Euclidean Patterns in Non-Euclidean Tilings). 3D euclidean networks are formed by a process of projection onto triply periodic minimal surfaces. Interpenetrated nets arise from 2D hyperbolic tilings by infinite tiles. A possible definition of 'knottedness' in 3D nets was proposed. A different view of nets is possible via sphere-packings graphs as shown by Elke Koch (Germany) with some exceptional cases including interpenetration of 3D nets, catenation of 2D layers and pairs of packings showing self-catenation. All the above apparently abstract results have been shown to exist with many examples taken from crystal structures that have been collected in a very useful web page by Stuart Batten (Australia). The experimentalist and crystallographer can find the tools to handle the complexity of all the above topologies in TOPOS as illustrated by Vladislav A. Blatov (Samara State U., Russia) where a complete classification of 3D nets has been illustrated by analyzing CSD and ICSD.

Davide M. Proserpio and Jean-Guillaume Eon

Non-ambient powder diffraction and kinetic studies

Peter Chupas (USA) gave an excellent overview of how rapid PDF information can be gained from samples using data collections as short as one second, opening up the possibility of studying both short and long range structure of materials as a function of time, temperature or during chemical reactions. Karena Chapman (Australia) illustrated the importance of considering both local and average structure in an excellent presentation on a variety of CN-bridged framework materials showing negative thermal expansion. Karena presented a comprehensive collection of data on a large family of these materials and gave clear insight as to the origins of this unusual effect. The importance of in-situ powder diffraction studies across a range of technological areas was nicely exemplified by the talks of Raj Suryanarayanan (USA) and Gilberto Artioli (Italy). Suryanarayanan showed how the kinetics of molecular polymorphic and desolvation reactions can be determined from lab data. These studies are of crucial importance for assessing the storage life of pharmaceuticals. He also showed how one can follow the crystallization of a material as a function of time and then deduce the amorphous-to-crystallize starting ratio, allowing quantitative analysis without internal standards. Artioli showed how in-situ studies allow one to follow important processes including the removal of template molecules from zeolites and hydrogen production from methane. René Guinebretière (France) described a new high temperature attachment for thin film diffraction studies and how its intrinsic thermal expansion can be automatically corrected to allow accurate determination of key sample microstructural parameters under non-ambient conditions.

Ivana Radosavljevic Evans

Structural biology of the immune system

This microsymposium reported recent advances in understanding the structure and function of molecules of the immune system, the line of defense of evolved organisms against pathogens. The number of high-resolution structures of antibody fragments, antigen-presenting molecules and their complexes with T-cell receptors, costimulatory molecules, and other proteins involved in this process is growing rapidly. These models are instrumental both for a more complete definition of the synergistic role of macromolecules in the immune response, and for opening new avenues for development of compounds with therepeutic potential. The session was opened by Ian Wilson (USA), who summarized the results deriving from the structural characterization of HIV-neutralizing antibodies, and the use of the antibody structures in a retro-vaccinology process. Of particular interest was the domain-swapped, carbohydrate-specific 2G12 antibody, that allowed the development of novel strategies for anti-HIV vaccines using rationally engineered antigens. Annette Shrive (UK) described the role of the Ca²⁺-dependent collectins in interacting with bacterial cell surface oligosaccharides, one of the mechanisms of antigen uptake by cells. Victor Streltsov (Australia) presented the structure of the unique IgNAR from the shark, the most primitive animal that has an adaptive immune system to fight infections. These antibodies differ from their mammalian counterparts since they are monomeric, and are substantially lacking a CDR2 region. Two structures of IgNARs, both unliganded and in complex with lysozyme, elucidate how these novel molecules can achieve the necessary hypervariability to interact with a broad spectrum of antigens. Michelle Dunstone (Australia) added another piece to the puzzle of the full-fledged, signaling-competent T cell receptor complex by presenting the structure of the CD3 εγ heterodimer in complex with the therapeutic antibody OKT3. The final talk from Piet Gros (The Netherlands) described the first structure of the C3 complement protein, the plasma protein with a central role in the cascade of events that leads to the formation of the membrane-attacking complex. The extremely complex, intertwined architecture of this member of the α2-macroglobulin superfamily of proteins is paralleled by a dramatic conformational change that occurs upon proteolytic activation.

Massimo Degano and E. Yvonne Jones

Structural knowledge and catalysis

This microsymposium focused on the relevance of structural knowledge in designing new catalysts, and studying their mechanism of operation. Agustin Galindo (Spain) described how oxy and thiodiacetate manganese complexes could be used as catalyst precursors in model oxidation reactions. Tsuneo Imamoto (Japan) lectured about high to almost perfect enantioselectivity using rhodium complexes and optically pure P-chiral diphosphine in some asymmetric hydrogenation reactions. He explained that enantioselectivity which is determined by the catalyst structure, also allows one to predict the mechanism of migratory insertion.

Antonio Mezzetti (Switzerland) addressed the issue of stereo selectivity in reactions of atom transfer in chiral ruthenium complexes, and stressed the role of molecular modelling studies and the need of crystallographic analysis to validate the method used. The theoretical point of view was emphasized by Walter Baratta (Italy) who spoke about agostic interactions studied with the help of X-ray and neutron diffraction techniques. He remarked that electron poor ruthenium (II) and platinum (II) complexes, stabilized by agostic M-η³CH₂ interactions, could be used as highly active transfer hydrogenation catalysts. Finally, Graeme Gainsford (New Zealand) presented recent advances in calculations based on Density Functional Theory (DFT) and attempts to predict the enantiomeric excess by means of those calculations.

J.C. Daran and P. Gómez-Sal

Time-resolved powder diffraction for materials production and processing

This session produced wide ranging descriptions of technique, and equally wide ranging, interpretations of their applications. Instrumentation included laboratory, neutron and synchrotron sources involving time resolutions ranging from seconds to hours. Another notable theme, was the wide range of sample environments needed, including several challenging experiments involving aspects such as 'one-shot' gas hydrate/clathrate formation, self-propagating reactions, single grain analysis during annealing/tensile stressing, high pressure (autoclave) acid-leaching, calcination/sintering up to 1500°C, and hydrogen gas-flow. The concluding message that emerged from this seesion is that the field has reached a stage that fast diffraction and in-situ capabilities are now the expectation and that current practitioners have to be continually ready to meet and cope with new requests.

Paul Barnes

Programming robust CIF and XML into crystallographic software

CIF is the Crystallographic Information File [www.iucr.org/iucr-top/cif/] which has become the standard for submission of crystallographic data sets of small molecules for archiving and publication. XML is the Extensible Markup Language that has become the standard for web-based data management. For many years these have seemed to be very distinct and incompatible frameworks for the management of data. The most important lesson of the microsymposium is that the differences are less important than the commonalities of the two frameworks and the complementary support these approaches offer to structural studies. In addition there was strong acceptance of an open source approach to software development, and the use of abstraction of data to aid in translation.

In the first talk, Peter Murray-Rust (UK) discussed conversion from CIF to XML in the form of CML (the Chemical Markup language) and the use of XML tools to process information from CIF data sets. To quote from the abstract, '…Open source and Open data provide a robust highthroughput crystallographic semantic web …'.
The theme of interoperability of CIF and XML was continued as Syd Hall (Australia) explained the importance of ontologies (dictionaries with relationships) in the management of data. Syd and Nick Spadaccini (Australia) gave a very convincing demonstration of a faithful translation between CIF and XML and explained the power of embedding validation logic into ontologies.
The third talk addressed some of the practical issues related to recent changes in the CIF format. G. Todorov (USA) began the talk, setting the general context and discussing some of the issues in moving smoothly between CIF and XML, such as the need to extend the character set used in CIF to agree with that in XML. Kostadin Mitev (USA) continued with a detailed explanation of a new open source utility that is being created for the IUCr publication process.
Joerg Kaercher (USA) discussed the automatic generation of data in multiple alternate formats from a database. The abstraction of the data as something distinct from its presentation is a very helpful way to move among formats and is one of the key concepts in easing the transitions among formats. In the last talk, Ganesh Jawahar Swaminathan (UK) extended the theme of database use to abstract data and showed its powerful use at EBI for support of the wwPDB and other initiatives requiring complex data management with data in multiple formats. In this case XML was the central format used. It was heartening to see an emerging consensus in an area that in past meetings has been contentious.

Herbert J. Bernstein and Brian McMahon

Electronic excitations

Electron Excitations can be measured with a number of probes. They can be probed with inelastic X-ray scattering (IXS) or X-ray Raman Spectra. The field is undergoing rapid progress due to experimental and theoretical advances and spawning, a growing number of workshops, symposia, and new theory networks, e.g., NanoEXC in the EU and the DOE Computational Materials Science Network in the USA.

The physics of electronic excitations goes beyond the ground state, one-electron properties of a material, as reviewed by W. Schuelke (Germany), a pioneer in the field. Schuelke noted that such excitations can be characterized in several equivalent ways, e.g., by the dynamical structure factor S, which is the Fourier transform of the densitycorrelation function, and is directly related to the fundamental dielectric response of a material. The same physical properties can also be measured with fast electrons using electron-energy loss spectra (EELS). The theory requires a quasi-particle description that goes beyond the independent particle approximation. Shuelke reported that a cumulant expansion model for the self-energy yields more accurate results.
Many body theoretical descriptions can be computationally intensive. Juha Soininen (Finland) reviewed detailed calculations of non renosant IXS (NRIXS) based on the (two-particle) Bethe-Salpeter equation and discussed a new multiplescattering approach, which has the potential to give much faster calculations over a wide energy range, and can be used to analyze NRIXS.
The nature of IXS can also be used to elucidate the many-body response of a system. John Hill (USA) reviewed how resonant IXS has been used to study the Mott-Gap, and the electronic excitations in transition metal oxides, Pedro Montano (USA) then discussed how magnetic IXS (e.g., magnetic Compton scattering) can be used to understand the physics of Manganites and to resolve the various theories responsible for large magnetoresistance in these materials. Finally T. Colson reviewed correlations in IXS using the low-loss plasmon dominated spectra.

The field of electronic excitations demonstrate that the synergy between theory and experiment have made it possible to interpret the experiments quantitatively and in terms of important dynamical physical properties of a system's response.

K. Hamalainen and J.J. Rehr

Crystallographic knowledge in drug design strategies

The aim of this microsymposium was to give an overview of the use of structural data at atomic resolution in the rational drug design of new molecules with therapeutic and diagnostic applications.

Ada Yonath (Israel) described how the analysis of high resolution structures of complexes of antibiotics with ribosomal particles has helped shed light on antibiotic selectivity and has provided a fair degree of understanding of some of the processes involved in antibiotic action (from reducing of decoding accuracy, via limiting conformational mobility, to interference with substrate binding and hindrance of the progression of growing proteins). Knowledge of such common principles is crucial and, when combined with genetic, structural and biochemical investigations, can be instrumental in structure-based approaches. She also suggested that combination treatment might be an efficient way to avoid drug resistance. Another interesting approach to overcome drug resistance was presented by Kalyan Das (USA). Structural and modelling studies of HIV-1 Reverse Transcriptase (RTs) in complex with potent dyaryl-pyrimidine nonnucleoside inhibitors showed that the conformational flexibility of these inhibitors appears to allow them to retain their potency against a wide range of drug resistant HIV-1 RTs, thereby providing a means to design drugs effective against rapidly mutating targets. Some of these newly discovered potent inhibitors are currently in clinical trial.

In computer-aided design, the limitations of crystal structures selected to perform docking experiments are often ignored and the reliability of the docking results can therefore be questionable. Interestingly enough, Noriaki Hirayama (Japan) showed, using the recently developed docking algorithm Ph4Dock, that Cruickshank's Diffraction-component Precision Index can be used to evaluate the reliability of the docking results.

Franck Leveiller and Michele Saviano

Advanced functional materials

This session focused on the interplay between structural features and physical and chemical properties in materials science with applications involving molecular crystals and inorganic oxides. The first presentation concerned electrical conductivity in ionic polymers and electronic metal complexes. Two important breakthroughs have been recently achieved: (i) the discovery of ionic conductivity in crystalline rather than amorphous polymer electrolytes, and (ii) the synthesis of molecular metals based on a single active molecule, rather than on two different interacting units. Yuri Andreev (UK) described the elucidation of the relationships between lithium ion mobility and ordered polymeric structure and methodological aspects of single-crystal and powder techniques for establishing the structure. Minoru Mitsumi (Japan) showed how semiconducting behavior in single-component molecular conductors may be coupled to trimerization or even hexamerization, depending upon the temperature of the active molecules within the linear chains building up the crystal structure. This highlights the details of the structural configuration of such important molecular materials.
Two talks focused on the emerging field of magneto-ferroelectrics. In these materials the frustration of parameters such as orbital ordering leads to complex spiral magnetic structures which break inversion symmetry and cause a ferroelectric lattice response. Tsuyoshi Kimura (USA) used this behavior as a search criterion for new magneto-ferroelectric materials and presented measurements illustrating the relationship between frustration, complex magnetic structure and ferroelectricity in CuFeO₂ and Ni₃V2O₈.
What is unusual and exciting about many of these materials is that the magnetic field couples strongly to ferroelectricity and can be used to manipulate the ferroelectric polarization as demonstrated recently for TbMnO₃. Nadir Aliouane (Germany) showed that the polarization flop in TbMnO₃ arises from a field induced incommensurate to commensurate magneto-structural transition. This transition is preceded by a change in the magneto-elastic coupling from quadratic (as suggested by spiral magnetic ordering) to linear with applied field. In fact the behaviour of these materials under magnetic field is rather complex, as highlighted by the devil's staircase like behaviour in TbMnO₃ shown in the figure.

Michele Catti and Dimitri Argyriou

Biomac OCM

Open commission meeting on biological macromolecules

Some years ago the Commission established clear guidelines for the submission of crystallographic data associated with the publication of macromolecular structures. The compliance within the community has been very good with structure factors now accompanying 90% of Protein Data Bank Depositions. The members of the Commission now felt that it was time to discuss the need to capture all available data associated with a protein structure determination. The underlying principle enunciated by Howard Einspahr (USA), Editor, Acta Crystallographica Section F, was the need to provide all data necessary for replication of the experiment in another laboratory. For a typical structure biology report, this might include details of the cloning, expression, purification and crystallization of the target protein. Zbyszek Otwinowski (USA) addressed the issue of which data needs to be deposited and, if so, in what format. He questioned the need to deposit all the computational steps along the way to a refined structure despite the ready availability of these data in most of the commonly used software packages. Tom Terwilliger (USA) illustrated how in macromolecular experiments, the final result does not always provide sufficient information about choices made in reaching the model. Harry Powell (UK) stressed the need for an audit trail and that it should be automatic and not an option. Data harvesting needs to commence at the start of the experiment. Howard Einspahr concluded the session with summary of the work he has been doing with the PDB and the Acta Crystallographica staff in Chester to expand the mmCIF dictionary to include all the new data items for the complete description of a structure analysis and the tools to extract these data in a form suitable for tabulation for presentation in a publication.

Mitchell Guss

COMCIFS OCM

Open commission meeting on COMCIFS

The strength of CIF lies in its dictionaries which provide precise descriptions of the ontology (i.e., the collection of concepts) that define crystallography. In the form approved by COMCIFS, these dictionaries can be found on the IUCr website. Their content can also be given in other representations such as the eXtensible Mark-up Language (XML). Zukang Feng (USA) described how the Research Collaboratory for Structural Biology (RCSB) in the US uses the mmCIF dictionary together with the CIF syntax to produce the archival form of the Protein Data Bank. RCSB receives macromolecular structure reports in a variety of formats which must be converted into CIF before they are added to the archive.

An XML version of the Protein Data Bank (PDBML) developed to assist data transfer between the partners collaborating in the World-Wide PDB was described by Haruki Nakamura (Japan). PDBML augments the information received from the Protein Data Bank with further biological and experimental details extracted from the literature with the intention of making PDB content available on-line with a variety of search and display options. Kim Henrick (UK) described how the Protein Data Bank is made available in the UK through the Macromolecular Structure Database.

Ralf Grosse-Kunstleve (USA), speaking from his experience in working with the Phenix project, pointed out that the lack of CIF software tools is the main reason many people prefer to work in XML: writing code for reading and writing CIFs is not a trivial exercise.

A solution to this problem was described by Heping Zheng (USA) who is constructing a CIF translation dictionary to provide the link between the internal memory of a program and the CIF that is to be written. To assist in the automatic production of publication-quality CIFs Zheng and his colleagues are adding routines to the HKL2000 suite of programs for macromolecular structure determination to produce a draft CIF and to allow the user to make final adjustments by hand. Ethan Merritt described a Python program library designed to import, manipulate and export macromolecular structure models written in CIF. He described the program TLSMD which analyzes the thermal motions of a protein, looking for the optimum way of breaking the molecule into rigid segments.

Brian McMahon (UK) described the use of CIF for reporting protein structures in the new Acta Crystallographica F. Structure reports of biological macromolecules are deposited in the usual way with the Protein Data Bank, which then provides the author with an mmCIF representation suitable for onward transmission to the journal where it is used for automatic generation of the experimental tables. Frank Allen (UK) reported that, in contrast to the Protein Data Bank, the Cambridge Structural Database now receives 96% of its structure reports in the form of CIFs, an achievement made possible because most small-molecule refinement programs are able to write their output in CIF format. The Cambridge Crystallographic Data Centre, which produces this database, uses CIF syntax internally and has recently released enCIFer, an editor and browser designed to allow people to create CIFs that conform to any given CIF dictionary.

I. David Brown.