Minutes of the Working Group 2 Meeting -------------------------------------- April 24th, Birkbeck -------------------- Present: Phil Evans (Chairman) MRC, Cambridge Jim Raftery Manchester Tadeusz Skarzynski (Industrial Rep.) Glaxo Wellcome Liz Potterton York/CCP4 Garib Murshudov York/CCP4 Misha Isupov Exter Jim Naismith St. Andrews Christine Cardin Reading Peter Keller EBI Martyn Winn Daresbury/CCP4 Alun Ashton Daresbury/CCP4 Adam Ralph Daresbury/CCP4 Cameron Dunn Bath Paul Holden Bristol Kim Henrick (Guest) EBI David Waller Sheffield Eleanor Dodson York Neil Isaacs (Chairman WG1) Glasgow Sheila Gover Oxford Ian Tickle Birkbeck Sue Bailey (CCP4 Secretary) Daresbury/CCP4 Lindsay Sawyer Edinburgh Apologies: Paul Emsley, Steve Yewdall 1. Minutes of Previous Meeting and Matters Arising. None. 2. Introduction to New Staff. The chairman introduced Alun Ashton to the committee. He position is at Daresbury and is on a two year contract. 3. Study Weekend 1998. The committee decided to confirm that the next Study Weekend on Databases would be at Reading on the 9th and 10th January 1998. There are to be three organisers Liz Potterton (from York), Judith Murray-Rust (from Birkbeck) and Ben Luisi (Cambridge). The preferred dates for the meeting made the choice of venue very limited. At the last meeting it was decided that two possibilities would be looked into, Reading and Sheffield. As Sheffield had no tiered lecture hall, Reading became the favourite. The lecture hall at Reading can only seat 400 at maximum, however there are side rooms that could be used as overflows if needed. These would be linked by video to the main hall. The accommodation is situated about 5 minutes walk away from the lecture hall. There are bar facilities and two dinning halls which can seat 400 and 200 people. Therefore, the conference dinner will need to be split, the same as last year. Also, afterwards a late bar will be provided in the SU. The organisers have already drawn up a list of possible action speakers. The numbers will need to be pruned for the meeting itself. organisers The committee was generally happy about the format of the meeting. There were just a few points that the committee felt were important. That there should be an explanation about why databases are important to crystallographers and what practical information can be gained by using them. That the format of the workshop should mirror the structure solution process, from crystallisation to refinement. There should be something on generating geometry dictionaries from databases and perhaps on something metal interaction databases. There was also some discussion on having computer demonstrations of the databases outside the main hall. The feeling was that this might draw people away from the lectures as free time will be limited. 4. Refinement Workshop, York Jim Naismith appealed for more people to apply for the Refinement Workshop, especially from Universities/Institutions which had not sent a participant. The one person per Institute rule maybe relaxed if not enough people apply. He stressed that there were a limited number of places available for the hands-on computer practicals but that more people could come to the lectures if they wished. Those interested should contact Jim Naismith. 5. Data Harvesting. There was a long discussion on whether CCP4 should back a action proposal from Kim Hendrick. Kim The basic idea is that some CCP4 programs would output specific data (such as R-factors, merging statistics etc.) that would be stored in a particular project directory. The data would be updated throughout the structure solution and would be used in assisting the user to deposite coordinates at the PDB. Also, the data would be stored at the EBI in a database designed by them. In order for this data harvesting to be successful, users must be aware that the data will be stored in a specific directory related to a project id. It therefore requires that a user be conscientious when running any CCP4 program to ensure that the data is stored correctly and to collate these files when outside their local computer environment. The format of these data files are to be based on ASN.1. This will make it easier for the EBI to assimilate the data into their database. These files are distinct from the working files used by CCP4 and are not intended to replace them. The EBI have said that they will undertake most of the coding associated with data harvesting and also provide any conversion routines between CCP4 formats and their database formats. 6. Recent Progress on REFMAC. Garib Murshudov gave a short talk about the progress made with REFMAC. There is a paper to be published in Acta Cryst. D part 3 about the use of anisotropic scaling and phased refinement of partial structures. Recent improvements to REFMAC include being able to apply an overall anisotropic B and holding the average overall B value constant during refinement. Finally, improvements have been made to bulk solvent parameters. 7. GUI Update. Liz Potterton talked about the progress so far with the GUI. The committee were shown the general layout strategy via a GUIlet to fft. More programs would need to be interfaced before a minimum GUI could be released to a wider community. Liz said that she would be considing various other aspects of the GUI in the next months. Job control by the GUI, a distinction could be made about how a job should be run i.e. interactively or in batch/background. Chaining 'tasks' which would be a first step to having more complex scripts run by the GUI. A task is defined as a operation like calculating a map via fft. However, tasks are not necessarily associated with one program, in the fft example, maps could be extended or converted to a different format. Reviewing results should be made easier with perhaps an index of log files. RasMol and a 2D viewer could be used as helper applications to display tables, maps, etc.. The addition of an AutoNotebook which would keep a log of the programs/tasks run and include user annotation. Finally, some strategy for removing intermediate files. It was suggested that a good sequel to the fft GUIlet action would be a interface to the solution of heavy atom sites. This Liz would quite complex and involve some of the points mentioned above. A new front end to RasMol has been designed in tcl/tk. This would facilitate communication between RasMol and the GUI. 8. Results from Finding Heavy Atoms in MAD Datasets. Unfortunately, Yao did not attend the meeting because it was felt that the meeting might overrun due to the other items on the agenda. However, Eleanor did ask if anybody had MAD data from structures with many heavy atom sites. This would provide a good test for Yao's work. 9. Program Suite. Changes in the next release will include the conversion of the documentation to HTML. The man pages will officially be extinct, although the doc files can be linked through cat1. Fred Vellieux wished to donate two programs OMIT (which action automatically combines omit maps to produce a 'complete omit DL map') and MFF2AT (which is similar to PEAKMAX). Currently, it is intended that OMIT should be included. Eleanor asked if the FHSCAL and SCALEIT could be merged. This might be difficult because of the different strategies. 10. Date of Next Meeting. Thursday 3rd July, at York. 11. AOB Pub! Adam Ralph, May 15th