Protein docking


Proteins often assemble into complexes to achieve a specific biological function. Obtaining high resolution atomistic structures of such complexes, however, is challenging. Software that can predict possible protein arrangements using available experimental data as guide can be exploited. The prediction of the structure of a protein assembly is however often difficult, because when interacting two binding partners may change conformation. These changes may be limited to local side chain rearrangements in the interacting surface, or may involve large scale rearrangements, which are difficult to predict. Further challenges arise from difficulties in properly combining diverse (and possibly inconsistent) analytical data. We:

  • integrate low resolution ensemble-based experimental data into the docking process handled by cutting-edge optimization techniques (see: POWer)
  • develop methods to appropriately score protein-protein electrostatic interactions (see: JabberDock)
  • develop methods to predict the assembly of multiple flexible protein copies according to arbitrary, deformable topologies