Meeting report

High throughput structures

[LA group] International panel of experts on supramolecular crystallography at the 2001 American Crystallographic meeting in Los Angeles include (first row) R. Tamura, Y. Ohashi, P. Scardi, M Olmstead,(back row) A. Beatty, J. Lipkowski, I. Huc, G. Gohel, N. Seeman, and M. Fujita.
The CCP4 meeting on 'High Throughput Structure Determination' (January, 2002, York), organized by R. Esnouf (Oxford) and K. Wilson (York), attracted nearly 500 participants.

T. Terwillinger (Los Alamos) described the NIH supported Tuberculosis Structural Genomics Consortium. The consortium has purified 156 of the 4000 gene products of the Mycobacterium tuberculosis, crystallized 40 proteins, and determined crystal structures of 22. In a session on high throughput, bioinformatics and optimal target selection, M. Weir (Inpharmatica, UK), described obtaining integrated and interpretable information. J. Gough (Cambridge) explained how the SUPERFAMILY web server ( can assign about half of all new structures to existing superfamilies and J. Peden made a pitch for the uses of Laboratory Information Management Systems (LIMS), P. Alzari (Pasteur Inst., Paris) described optimization of target selection; genomic analysis; functional analysis; proteomics; and comparative genomics. E. de la Fortelle described LIMS that are operational at Structural GenomiX in San Diego. The software tracks highly-automated experiments from the cloning stages through protein expression and purification, crystallization, data collection, structure solution, and 'publication' in the SGX Virtual Journal. Membrane-bound targets are being examined using automated detergent screens. Robotic crystallization for storage/retrieval and image capture allow examination of 1000 plates a day.

A session on techniques included descriptions of recombinant protein expression (I. Jones, Reading), mutant E. coli hosts able to overproduce high levels of globular and membrane proteins (F. Hill, Avidis), 96-well plates for the vapor diffusion methods, a multichannel-pipetting robot for drop preparation, a detection system for identification of crystals (L. Nyarsik, Berlin Protein Structure Factory), high-throughput structure determination by NMR (I. Campbell, Oxford) and electron crystallography of very small crystals (J. P. Abrahams).

In a session entitled 'Crystallization/Data Collection' G. Spraggon (GNF) and J. Wilson (York U.) described systems that scan automatically to detect crystals, M. Walsh (ESRF-UK) described robotic devices for mounting crystals developed at ALS, SSRL, APS, and by MSC/Rigaku and A. Thompson (EMBL/ESRF) described developments in beamline automation at ESRF, A. Leslie (Cambridge) reviewed the needs of an 'expert system' to reduce the whole process to pressing a button. We may soon be sending crystals to the synchrotron in steel blocks in a transport Dewar with bar codes. In a session on phasing and software, M. Winn outlined ongoing developments in the CCP4 program package, Y. Jia-Xing described ab-initio structure determination of proteins for which atomic resolution data is available with the program ACORN, G. Murshudov talked about recent advances in the REFMAC maximum likelihood refinement program and P. Adams concluded the session with a description of the PHENIX software initiative.

With respect to molecular graphics, D. Turk presented improvements to his Main program for automatic building of maps even at moderate resolution (, L. Potterton presented the plans for the CCP4 viewer and demonstrated it in its current rasmol-like state and A. Jones discussed the latest bells and whistles on O ( C. Nave reminded us that there are cultural changes on the way in macromolecular crystallography; 'you may be worried by this or excited by it but it will happen'. The volume of the data that worldwide efforts in high-throughput structure determination will yield is daunting and downstream analysis may prove to be the 'real bottleneck' in the future.

BCA Crystallography News, No. 80