Poster #RP103
Improved tools for coiled-coil identification, analysis and prediction
John Walshaw*, Efrosini Moutevelis**, Oli Testa**, Derek N. Woolfson**
*John Innes Centre, Norwich, UK; **University of Bristol, Bristol, UK
The coiled coil is one of the most common and regular protein-structural motifs. It consists of 2 or more alpha-helices packed together via a characteristic interlacing of their side chains (knobs-into-holes). In turn this is based upon a 7-residue pattern (the heptad). Successful sequence analysis methods for predicting the presence of coiled coils have long been available. However, challenges remain in predicting binding partners and other aspects of coiled-coil assembly and design. Previously we reported an algorithm, SOCKET, that detects the characteristic core-packing interactions in three-dimensional structures. We present an improved version of this software, which highlights interruptions in the repeats, and the ability to correctly interpret modified amino acids. Secondly, SOCKET has enabled the collection of statistics concerning the choices of amino acid in each heptad position preferred by by different coiled-coil topologies: 5.5% of structures in the Protein Data Bank have coiled-coil motifs; 87% are dimeric, 10% trimeric, 3% tetrameric. We also describe more complex, higher-order assemblies mediated by the same fundamental packing interactions. These data have implications for the prediction, from sequence, of topologies and binding partners of putative coiled-coil helices. With this goal in mind, we introduce the development of a new, interactive relational database, CCPLUS.