The organizers of the blind-test for docking methodologies called CAPRI have just announced the start of round 12. The target is a large protein, of more than 300 amino-acid residues, and the information we got from the organizers is that it forms a symmetric homo-dimer. That means that two copies of the same protein interact with one another in a way that rotating one copy around an axis of symmetry, produces the second copy.

Such 'symmetric' complexes are quite common in molecular biology, and have been observed, e.g., in virus capsids (the protein envelope of the virus) and in amyloid fibers, which have been associated with Kreutzfeldt-Jacob and Alzheimer's diseases. These assemblies are often very large, potentially consisting of tens of protein copies, but symmetry allows us to make a simplifying assumption that all copies of the molecule adopt the same internal conformation. So, using the symmetry assumption, we can in fact restrict the conformational space of the problem significantly, and model these very large complexes.

Because symmetric assemblies are important for understanding molecular biology, and can be treated computationally, two members of the Baker group, Ingemar Andre and Phil Bradley, have recently developed a method to incorporate symmetry into docking, and put this new capability in BOINC. The current CAPRI target provides an excellent platform for testing their new developments. Over the next few weeks, you will see jobs entitled 'CAPRI_12' that will test various starting positions and assumptions regarding the complex. With the large number of decoys that we will get from these BOINC runs, we hope to identify the native structure.

---