Accurate Protein Docking by Shape Complementarity Alone

Elucidating the molecular details of protein-protein interactions is essential to understanding cellular processes. Given the recent increase in protein structural information, mostly of monomeric proteins, we now have the data necessary to address protein-protein interactions by computational approaches. Previous attempts at in silico protein docking generate a multitude of answers that have similarly high scores for both correctly and incorrectly docked proteins. We have developed the first algorithm that successfully predicts the re-docking of known protein-protein complexes without any false positives. Because our algorithm is based on complementarity alone, this implies that shape matching suffices for recognizing correct docking configurations. The essential features needed to achieve accurate protein docking are a fine covering of the space of rigid motions and a score function that counts the number of atoms at close distance. Our results provide a proof-of-principle for the development of faster docking methods based on shape complementarity alone that incorporate protein flexibility.