Dockground resource for protein recognition studies
Deepak Singla, Petras Kundrotas, Ilya Vakser, Computational Biology, University of Kansas
Structural characterization of protein–protein interactions (PPI) is essential for understanding life processes at the molecular level. Computational approaches to structure prediction are important because of the large gap between known protein sequences and experimentally determined protein structures. Development of the computational approaches requires sets of reference structures for modeling, validation, and benchmarking. Dockground is a comprehensive resource for protein recognition studies, containing such pre-compiled protein-protein sets for various purposes, as well as user-friendly interfaces for generating sets according to specific user requirements. The current version contains 298,287 pairwise complexes from 103,603 PDB Biounit entries. It includes precompiled protein-protein unbound X-ray docking benchmark sets, simulated unbound structures sets, docking decoys (scoring benchmark), docking benchmark sets of protein models, and template sets for docking by structure alignments. The user interface allows filtering of the data according to the presence of ligands at the interface, di-sulphide bonds, and alternative binding modes, and removal of redundancies based on sequence or structural similarities. The resource is available at http://dockground.compbio.ku.edu.