DrugScorePPI Knowledge-Based Potentials Used as Scoring and Objective Function in Protein-Protein Docking

Dennis M Krüger,Holger Gohlke,Pablo Chacón,José Ignacio Garzón,Paolo Carloni

doi:10.1371/journal.pone.0089466

Abstract

The distance-dependent knowledge-based DrugScorePPI potentials, previously developed for in silico alanine scanning and hot spot prediction on given structures of protein-protein complexes, are evaluated as a scoring and objective function for the structure prediction of protein-protein complexes. When applied for ranking “unbound perturbation” (“unbound docking”) decoys generated by Baker and coworkers a 4-fold (1.5-fold) enrichment of acceptable docking solutions in the top ranks compared to a random selection is found. When applied as an objective function in FRODOCK for bound protein-protein docking on 97 complexes of the ZDOCK benchmark 3.0, DrugScorePPI/FRODOCK finds up to 10% (15%) more high accuracy solutions in the top 1 (top 10) predictions than the original FRODOCK implementation. When used as an objective function for global unbound protein-protein docking, fair docking success rates are obtained, which improve by ∼2-fold to 18% (58%) for an at least acceptable solution in the top 10 (top 100) predictions when performing knowledge-driven unbound docking. This suggests that DrugScorePPI balances well several different types of interactions important for protein-protein recognition. The results are discussed in view of the influence of crystal packing and the type of protein-protein complex docked. Finally, a simple criterion is provided with which to estimate a priori if unbound docking with DrugScorePPI/FRODOCK will be successful.

Highlights

Protein-protein interactions have important implications in most complex cellular signalling processes [1]
Less than 500 interactions were found between N.3 and positively charged atomtypes as well as for S.3 and positively or negatively charged atomtypes. Such interactions are rather unlikely to occur when evaluating proteinprotein complex configurations, too, and should not grossly affect the scoring results. These results indicate that the knowledge base of 851 protein-protein complexes for derivation of DrugScorePPI is large enough to yield statistically significant potentials despite the smaller number of complexes used for deriving DrugScorePPI than for deriving DrugScore [38]
When compared to the probabilities for random selection, DrugScorePPI shows superior performance in ranking acceptable solutions on the top. Comparing these results to the ones of Baker and coworkers (Table 1) shows that DrugScorePPI performs slightly inferior in the case of the ‘‘unbound perturbation’’ dataset but superior in the case of the ‘‘unbound docking’’ dataset

Summary

Introduction

Protein-protein interactions have important implications in most complex cellular signalling processes [1]. Several studies pointed to the existence of hotspot residues that account for most of the binding free energy in these interfaces [4,5,6,7,8]. These hotspots help guiding the development of modulators of protein-protein interactions [9]. Compared to the case of single protein structures, the number of experimentally determined structures of protein-protein complexes is still very limited To overcome this limitation, various protein-protein docking approaches have been developed for predicting the structure of protein-protein complexes [12,13,14,15,16,17]

Methods

Results

Conclusion