Binding site prediction and improved scoring during flexible protein-protein docking with ATTRACT

Sébastien Fiorucci, Martin Zacharias

Research output: Contribution to journalArticlepeer-review

58 Scopus citations


The ATTRACT protein-protein docking program combined with a coarse-grained protein model has been used to predict protein-protein complex structures in CAPRI rounds 13-19. For six targets acceptable or better quality solutions have been submitted (high quality predictions for targets 32, 40, 41, and 42). The improved performance compared to previous rounds can be attributed in part to the inclusion of conformational flexibility during systematic searches and an optimized scoring function. In addition, a recently developed method for the prediction of putative protein binding sites based on the electrostatic penalty to place neutral low dielectric probes on the protein surface was applied to the most recent targets. The approach resulted in useful predictions of putative binding sites that can help to limit the systematic docking searches. Possible improvements of the docking approach in particular at the scoring and refinement steps are discussed.

Original languageEnglish
Pages (from-to)3131-3139
Number of pages9
JournalProteins: Structure, Function and Bioinformatics
Issue number15
StatePublished - 15 Nov 2010
Externally publishedYes


  • Conformational change
  • Desolvation
  • Docking prediction
  • Elastic network model
  • Electrostatic interaction
  • Normal mode analysis
  • Protein-protein docking


Dive into the research topics of 'Binding site prediction and improved scoring during flexible protein-protein docking with ATTRACT'. Together they form a unique fingerprint.

Cite this