Newsletter contents... UP

CCP4/Max-INF Workshop on Refinement and Validation of Macromolecular Structure

Wednesday 3rd January to Tuesday 9th January, University of York

Organisers: Garib Murshudov and Eleanor Dodson (York)


Short courses have proved to be a valuable way to teach crystallographic techniques. This workshop aimed to provide a sound training in refinement techniques as part of the avowed CCP4 and EU funded MAX-INF initiatives to promote good practice. Such workshops also give valuable feed-back for method developers, and helps them test their programs against a wide variety of data sets and models.

Format of the course

The course was divided into morning lectures and afternoon tutorials where students were introduced to the software, and encouraged to apply it to their own data. (The full timetable is available at

The University of York was able to provide two computer classrooms each able to accommodate twenty students, so with some apprehension we accepted 40 participants, who were divided into two tutorial groups. This meant each tutor had to repeat the class twice, and that they had less time with individual students, but it did allow members of more laboratories to have access to this very intensive program.

Course material can be accessed from the above URLs listed in the program.

Details of the course

The tutors were :

Paul Adams CCI Lawrence Berkeley Laboratory
Ralf Grosse-Kunstleve CCI Lawrence Berkeley Laboratory
Gerard Bricogne Global Phasing
Pietro Roversi Global Phasing
Eric Blanc Global Phasing
Richard Morris Global Phasing
Garib Murshudov University of York & CCP4, CLRC
Eleanor Dodson University of York
Liz Potterton University of York
Martyn Winn CCP4, CLRC
Victor Lamzin EMBL Hamburg
Anastassis Perrakis NCI Amsterdam
Randy Read Cambridge
George SheldrickUniversity of Gottingen
Dale Tronrud Eugene, Oregon

The software packages used were: CNS, BUSTER, REFMAC and associated CCP4 software, ARP/wARP, SHELX, and TNT.

The standard of the lectures was excellent; as a community we are very privileged that the developers are willing to devote so much time and effort to teaching. In particular Paul Adams and Dale Tronrud both came from the West Coast of USA, and endured jet lag and British weather for the dubious satisfaction of working 12 hour days, for the reward of free institutional meals. George Sheldrick presented the fundamentals of refinement as well as providing excellent tutorials in the use of SHELX. It was very helpful to have such a clear exposition of this material early in the course and all other lecturers benefited from and built on his presentations. It was pointed out that he was one of the few lecturers who is a full-time academic and although this must limit his time for research it is a great boon for his teaching style.

As well as the listed lectures there were several valuable discussions on special topics. One of these was scaling the calculated and observed structure factors when the model is incomplete (always true to some extent). How best to parameterise solvent, and the unmodelled parts of the unit cell is still a matter for research and it was valuable for the developers to exchange experience and I hope interesting for the students to listen to the discussion and realise that there is often not a "correct" answer to these problems.

Another centred on how to describe and deposit the geometric and stereochemical restraints used during refinement. Traditionally many of these criteria were used as the basis of validation, but obviously this is not appropriate when they are also restrained. Kim Henrick from the European Bioinformatics Centre described the way they plan to both store the target values and report on individual structures.

As usual the end-of-course party was a great occasion, as illustrated by some candid camera shots! Pietro and Eric are welcome ANY TIME they wish to come to York; they kept up a stream of clean crockery and cutlery from our overcrowded kitchen.

Problems of the course

The students were too polite to complain about the overcrowding, and I did not hand out a questionnaire to get such feedback. The classrooms were overcrowded, and the computing resources stretched to their limit. Probably conditions would have been more comfortable with fewer students, but choosing between people from different Universities was invidious and some people made late but compelling cases for inclusion. In the end we felt it better to have 40 people 80% satisfied rather than 20 people 90% satisfied.

An additional problem with this course was travel to and from York. The course was planned before the disastrous floods which washed away part of the York-London railway track. However everyone arrived in the end, despite our anxieties.


Garib and Eleanor are enormously grateful to all those who spoke and demonstrated, without them there would have been no course. Kevin Cowtan with the CCP4 staff, in particular Liz Potterton and Alun Ashton, worked extremely hard to guarantee that the network functioned, and that the software was properly installed and accessible, a formidable task. The computing staff at York were very helpful and made sure things went smoother than they should have done. The students worked extremely hard and were very tolerant of defects in the organisation. No student/post-doc had to pay a penny to come thanks to CCP4 and the EU initiative.

Newsletter contents... UP