Functional Proteomics

Volume 484 of the series Methods in Molecular Biology pp 505-521

Prism: Protein-Protein Interaction Prediction by Structural Matching

  • Ozlem KeskinAffiliated withCenter for Computational Biology and Bioinformatics and College of Engineering, Koc University
  • , Ruth NussinovAffiliated withBasic Research Program, SAIC-Frederick, Inc. Center for Cancer Research Nanobiology Program NCI-FrederickSackler Institute of Molecular Medicine, Department of Human Genetics and Molecular Medicine, Sackler School of Medicine, Tel Aviv University
  • , Attila GursoyAffiliated withCenter for Computational Biology and Bioinformatics and College of Engineering, Koc University

* Final gross prices may vary according to local VAT.

Get Access


Prism (protein interactions by structural matching) is a system that employs a novel prediction algorithm for protein-protein interactions. It adopts a bottom-up approach that combines structure and sequence conservation in protein interfaces. The algorithm seeks possible binary interactions between proteins through structure similarity and evolutionary conservation of known interfaces. It is composed of a database containing protein interface structures derived from the Protein Data Bank (PDB) and predicted protein-protein interactions. It also provides related information about the proteins and an interactive protein interface viewer. In the current version, 3799 structurally nonredundant interfaces are used to predict the interactions among 6170 proteins. A substantial number of interactions are verified in two publicly available interaction databases (DIP and BIND). As the verified interactions demonstrate the suitability of our approach, unverified ones may point to undiscovered interactions. Prism can be accessed through a user-friendly website ( and it will be updated regularly as new protein structures become available in the PDB. Users may browse through the nonredundant dataset of representative interfaces on which the prediction algorithm depends, retrieve the list of structures similar to these interfaces, or see the results of interaction predictions for a particular protein. Another service provided is the interactive prediction. This is done by running the algorithm for the user input structures.

Key Words

Protein interactions protein interaction prediction protein interfaces protein databases