RosettaDock has been increasingly used in protein docking and design strategies in order to predict the structure of protein-protein interfaces. Recently, we implemented RosettaDock in an object-oriented fashion as part of the newly developed Rosetta v3.1 modeling suite. The new version of RosettaDock includes minor algorithmic changes as well as the introduction of greater code flexibility and extensibility, allowing us to model a wider variety of protein-protein interfaces and additional interface molecules that are important in biologically relevant interactions. Here we test the capabilities of RosettaDock v3.1 against the expanded Docking Benchmark 3.0, and compare it with RosettaDock v2.3, the latest version of the previous Rosetta software package. The expanded benchmark contains a diverse set of 115 docking targets including 22 antibody-antigen complexes, 33 enzyme-inhibitor complexes, and 60 ‘other’ complexes. These targets were further classified by expected docking difficulty into 84 rigid-body targets, 17 medium targets, 14 difficult targets based. We carried out local docking perturbations for each docking target, using the unbound structures when available, in both RosettaDock v2.3 and v3.1. We defined a docking funnel as any docking result in which at least three of the five lowest energy structures were within at least 5A interface RMSD from the crystal structure of the complex. Overall the performances of RosettaDock v2.3 and v3.1 compared similarly to each other. RosettaDock 3.1 achieved 49 docking funnels, compared to 48 in v2.3. In terms of complex-type, RosettaDock v3.1 achieved docking funnels for 50% of antibody-antigen targets, 46% of enzyme-inhibitor targets, and 30% of ‘other’ targets. In terms of docking difficulty, RosettaDock v3.1 achieved funnels for 45% of rigid-body targets, 30% of medium targets, and 14% of difficult targets. Additionally, we report new capabilities in RosettaDock v3.1 to incorporate non-protein moieties such as small-molecule ligands, cofactors, and post-translational modifications in order to more faithfully model biologically relevant protein-protein interfaces. This study marks the most extensive benchmarking of the RosettaDock module to date and establishes a baseline for future research in protein interface modeling and structure prediction.
This preliminary program is subject to change.
![[ Visit Client Website ]](images/banner.gif)
Improvements to Computational Protein-Protein Docking Methods Allow for the Modeling of a Wider Variety of Biologically Relevant Protein-Protein Interfaces
RosettaDock has been increasingly used in protein docking and design strategies in order to predict the structure of protein-protein interfaces. Recently, we implemented RosettaDock in an object-oriented fashion as part of the newly developed Rosetta v3.1 modeling suite. The new version of RosettaDock includes minor algorithmic changes as well as the introduction of greater code flexibility and extensibility, allowing us to model a wider variety of protein-protein interfaces and additional interface molecules that are important in biologically relevant interactions. Here we test the capabilities of RosettaDock v3.1 against the expanded Docking Benchmark 3.0, and compare it with RosettaDock v2.3, the latest version of the previous Rosetta software package. The expanded benchmark contains a diverse set of 115 docking targets including 22 antibody-antigen complexes, 33 enzyme-inhibitor complexes, and 60 ‘other’ complexes. These targets were further classified by expected docking difficulty into 84 rigid-body targets, 17 medium targets, 14 difficult targets based. We carried out local docking perturbations for each docking target, using the unbound structures when available, in both RosettaDock v2.3 and v3.1. We defined a docking funnel as any docking result in which at least three of the five lowest energy structures were within at least 5A interface RMSD from the crystal structure of the complex. Overall the performances of RosettaDock v2.3 and v3.1 compared similarly to each other. RosettaDock 3.1 achieved 49 docking funnels, compared to 48 in v2.3. In terms of complex-type, RosettaDock v3.1 achieved docking funnels for 50% of antibody-antigen targets, 46% of enzyme-inhibitor targets, and 30% of ‘other’ targets. In terms of docking difficulty, RosettaDock v3.1 achieved funnels for 45% of rigid-body targets, 30% of medium targets, and 14% of difficult targets. Additionally, we report new capabilities in RosettaDock v3.1 to incorporate non-protein moieties such as small-molecule ligands, cofactors, and post-translational modifications in order to more faithfully model biologically relevant protein-protein interfaces. This study marks the most extensive benchmarking of the RosettaDock module to date and establishes a baseline for future research in protein interface modeling and structure prediction.