Abstract

Protein-Protein Interactions (PPIs) play a very important role in many cellular processes and a variety of experimental approaches have been developed for their identification. These approaches however are partial, timeconsuming and they usually suffer from high error rates. Recently, computational methods have been employed to assist for the prediction producing encouraging results. With this work we offer a critical review of recent computational PPI prediction methods by evaluating their strengths and limitations. Moreover we discuss open problems common to all schemes and try to suggest solutions. Finally we propose future research directions which could potentially more effectively handle some of the restrictions of existing approaches.

Keywords: Protein-Protein interactions, computational methods, machine learning, databases, experimental methods, SVM, random forests, Bayesian Classifiers, Neural Networks