Optimizing a coarse-grained model for the recognition of protein-protein binding

Abstract

We are optimizing a force-field to be used with our coarsegrained protein model for the recognition of protein -protein binding. We have found that, apart from ranking correctly the ligand-receptor conformations generated in a protein-protein docking algorithm, our model is able to distinguish binding (experimental structure) from nonbinding (false positive) conformations for many complexes. This suggests us that the model could have a good performance in complete cross-docking, a method aimed to recognize the possible binding between any two proteins that are unknown to interact.