Minutes of the Working Group 2 Meeting -------------------------------------- Wednesday Jan 27th 1999, University of York -------------------------------------------- Present: Phil Evans (Chairman) MRC, Cambridge Martyn Winn Daresbury/CCP4 Liz Potterton York Peter Briggs Daresbury/CCP4 Fred Vellieux IBS/Grenoble Jim Naismith St. Andrews Sue Bailey Daresbury/CCP4 Christine Cardin Reading Eleanor Dodson York Paul Emsley Glasgow Kevin Cowtan York Tadeus Skarzynski Glaxo Wellcome Jan White Sheffield Alun Ashton Daresbury/CCP4 Sheila Gover Oxford Jorge Navaza Ian Tickle Birbeck Garib Murshudov York Alexei Vagin York Chris Gilmore Glasgow Apologies: Peter Keller, Kim Henrick, Peter Moody, Kay Wilkinson, Harry Powell, Misha Isupov, Cameron Dunn 1) Minutes of meeting of 15-9-98 and matters arising The minutes were accepted and there were no matters arising. 2) Study Weekend '00 Date - Jan 7/8 Venue - York Topic - ??? Organisers - ??? There was much discussion over the topic and organisers for the next Study Weekend. Sue reported that the topic of "Low Resolution Phasing" had been suggested by WG1. The consensus was to bring together several topics: Low Resolution Phasing, Electron Microscopy (EM), with one or more of: Ab Initio Phasing Methods, Molecular Replacement, Phase extension methods. The possibility that such a range of topics might lack cohesion prompted the suggestion that each topic be treated in a separate session. Helen Saibil (as a representative of the EM community) and Julie Wilson (Ab Initio methods) were suggested as organisers. It was felt that a third organiser was required, with Sheila Gover suggesting that she ask Jonathan Grimes. Andrew Leslie and Per Bullough could be asked if he was unavailable. It was thought that there wouldn't be serious problems with communications between three people. Sue pointed out that organisers now also have more work after the Study Weekend due to Acta Crys. publishing the proceedings. Sheila suggested that e-mail addresses rather than the full institution postal addresses be printed in the participant lists, following comments from the Sheffield meeting. It was agreed that this measure would both save paper and also prove generally more useful for communications, and so will be adopted in future. It was generally agreed that the feedback from this year's Study Weekend had been extremely positive. ACTION: Sue to approach potential Study Weekend organisers 3) Workshops IUCr99 and Computing School: Martyn asked about CCP4 involvement with the IUCr Computing School at Hinxton Hall. At present CCP4 is only providing funding and further involvement seems unlikely. Concerns were voiced about the cost of the Hinxton Hall conference facilities but Chris Gilmore made assurances that this would not be prohibitive. Ideas for further CCP4 workshops: It was suggested that a CCP4 workshop should be tagged on the opening day of the IUCr, re-running the Prague workshop for around 100 people. Paul Emsley agreed to act as the local organiser. Chris Gilmore urged that action needs to be taken quickly, for the booking of lecture space and other facilities. CCP4 will pay the bulk of the costs but it was decided that a nominal fee should be charged to encourage those who registered to turn up. Sue highlighted the difficulties involved in processing small registration fees through CLRC and alternative solutions were suggested. The Executive Committee needs to approve funding of the workshop. There was a call for more people to organise workshops along the lines of the Bath Summer School or the Refinement Workshop. Students find the workshops immensely useful but the amount of work involved means they cannot be run very often unless more people are prepared to act as organisers. A list of possible speakers/tutors (with topics) is being compiled by WG1 as a resource for future workshops. Ian Tickle commented that the lack of facilities (particularly computers) is also a problem for some venues. The issue of a portable CCP4 machine for workshop demonstrations was also raised. It was agreed that the computer should be purchased as soon as possible. ACTION: CCP4 staff to contact Chris Gilmore re: workshop, Paul Emsley 4) Review of imminent release of 3.5 Martyn reported that CCP4 release 3.5 should be made by the end of February. Although this will not include the GUI or CIF, there are enough accumulated changes to justify a new release. There is also a need to achieve a stable release 3.5 for a separate GUI release. It is intended that later releases of the GUI will be synchronised with the rest of the suite. The new release will include: changes for data harvesting; Kevin Cowtan's library for generating html logfiles; new programs DYNDOM, MTZMADMOD, DETWIN, REFMAC with anisotropic refinement, TOP, WATNCS and FINDNCS; automated X-windows installation. There are also other minor changes. ACTION: Martyn 5) New HTML (with applets) log files Kevin Cowtan gave a brief overview of HTML logfiles from CCP4 programs, and new library routines (libhtml) developed to achieve this. These include a java applet which can display log-graphs graphically through a HTML browser. The addition of HTML can enhance logfile clarity whilst maintaining backward compatibility, so that normal methods of viewing the logfile content are still applicable. It was agreed that over a period of time the other programs should also be upgraded to produce HTML-ised logfiles. Liz commented that these changes should be accompanied by a review of logfile contents, and that XML tags would in the future make it easier to extract different levels of information from the logfile. 6) GUI update Liz gave a summary of the progress made with the GUI since the last meeting. There are a number of "alpha test sites" currently using the GUI but Liz would like more testing and feedback, particularly for the installation of Tcl/Tk (which the GUI needs in order to run). She plans to tour UK sites to give demos of the GUI, and asked for people to host these demonstrations. She also commented that work needed to be done on the documentation, in particular the tutorials, and she asked for help with this and with documenting loggraph output. Another issue was the problem of keeping the GUI up-to-date with changes in the programs. Liz commented that a mechanism is needed to record when keywords are changed in order to facilitate this process. 7) Closer integration of CCP4 with data collection/collaboration with Daresbury staff Martyn distributed a handout in which Colin Nave outlined ideas for greater integration of the data collection process and the CCP4 programs, to ease the user's transition from one stage to the other. Existing systems for interfacing data collection e.g. DPS were felt to be limited and there is room for a more general interface. This would probably be an extension of the CCP4 GUI, with the aim initially being an interface for MOSFLM and SCALA developed by someone working at a synchrotron. Phil reported that this idea has the support of WG1, and that a new person could be hired to do this, possibly a joint CCP4/Daresbury-PX funded post. A tangential discussion ensued regarding the need for a standard format for image file headers which detector buyers can request from manufacturers. Phil agreed to write and distribute a specification of the ImageCIF format for this purpose, and will contact Bob Sweet for a status report. ACTION: Martyn, Phil (has since contacted Bob Sweet) 8) Status of NT port Jim Naismith argued that there is an urgent need for a CCP4 port to NT, now that PCs running NT have become viable alternatives to UNIX machines, and that this would be a full time job for whoever was employed to do it. Tendering out to a commercial firm would be too expensive and make later updates difficult for CCP4 staff. Alun explained that there are issues with binary files and with program stability under NT, with further complications due to the different flavours of NT. Commercial compilers are expensive for users but this could be circumvented by distributing executable code (like e.g. CNS). This idea was popular. A port of the GUI will also be required - Liz reported that Tcl/Tk should run under NT, but BLT (used for loggraph) may be more problematical. It was suggested that an advert on the bulletin board might attract someone to attempt the port. Discussions should be held with Joseph Krahn (who has ported the x-windows software is trying the GUI) and Alwyn Jones (who has had some success trying to port O). There was also the possibility that the new DL post might be used to work on the NT port. ACTION: Alun 9) Progress with REFMAC Garib confirmed that REFMAC with anisotropic refinement would be available in the next release of CCP4 along with a beta release with mixed refinement (iso + aniso). The next step is incorporating TLS parameters with individual anisotropic refinement, in collaboration with Martyn. Initial results are promising but more testing is required. Garib reported that the replacement for PROTIN is now his top priority. ACTION: Garib, Martyn 10) Progress on MOSFLM Phil read a short report from Harry Powell: MOSFLM now has autoindexing; there are some problems building on IRIX6.5 (mixing of 32/64 bit objects which is also a problem with other programs) and running on Alphas (difficulties with setting system resources appropriately); there is now a prototype FAQ webpage for MOSFLM; tests comparing the speeds of DENZO and MOSFLM suggested there was little difference in running times. 11) Date of next meeting Wednesday 5th May 1999 at Birbeck College London. 12) AOB Jorge Navaza has donated a recent version of AMoRe to CCP4 which will need work to bring into line with the CCP4 conventions.