Docking in CAPRI

This issue of Proteins takes you to the Gulf of Naples, with a bright blue sky above the Mediterranean sea and Mount Vesuvius in the background. Soon we shall be landing on the Isle of Capri, famous for its cliffs, its long history, and all the artists, writers, and poets who visited or lived there. Or shall we? Well, not quite. Today’s CAPRI is not a sea resort, but the Critical Assessment of PRedicted Interaction experiment, and we are docking proteins instead of a yacht. The CAPRI experiment aims to test the capacity of docking methods: can we predict the structure of a protein–protein complex if we know that of its components? Like protein fold predictions in CASP (Critical Assessment of methods for Structure Prediction), a similar experiment which is well known to the readers of Proteins, a CAPRI prediction must be made blindly before being compared by independent assessors to an unpublished experimental structure. All groups worldwide that develop methods to analyze protein–protein interaction and predict the structure of complexes are invited to participate. Although CASP is 10 years old, the CAPRI experiment began in 2001, and two rounds of predictions were held in July 2001 and January 2002. A third round began in January 2003; others will follow provided suitable targets are found. Here we report results obtained during the first two rounds of CAPRI by some 20 groups making predictions on seven targets: seven protein– protein complexes for which crystallographers had made new X-ray structures available to us. R. Mendez, S. Wodak, and collaborators in Brussels had the difficult task of assessing the predictions against the crystallographic atomic coordinates. Kim Henrick set up and maintained the CAPRI Web site (http://capri.ebi.ac.uk) at the European Bioinformatics Institute of Hinxton (UK), where all predictions were submitted and preliminary results disclosed. The experiment relied first on the generosity of my colleagues in Gif-sur-Yvette and Marseille (France) and in Rockville (MD) who provided the X-ray structures of the targets, then on the expertise and sagacity of the predictor groups, and finally on the hard work of the assessors and the CAPRI webmaster. Whether rounds 1 or 2 were successful, you will judge by touring CAPRI by means of its Web site and of the reports presented here. We hope the tour will be inspiring and that it will entice the visitors to participate in future rounds of the experiment by providing targets or by making predictions.