Computational Methods for Prediction of Protein-Protein Interactions: PPI Prediction Methods

Computational Methods for Prediction of Protein-Protein Interactions: PPI Prediction Methods

Sneha Rai (Netaji Subhas Institute of Technology, India) and Sonika Bhatnagar (Netaji Subhas Institute of Technology, India)
Copyright: © 2017 |Pages: 32
DOI: 10.4018/978-1-5225-1798-6.ch012

Abstract

The key signaling pathways in cellular processes involve protein-protein interactions (PPIs). A perturbation in the balance of PPIs occurs in various pathophysiological processes. There are a large numbers of experimental methods for detection of PPIs. However, experimental PPI determination is time consuming, expensive, error prone and does not effectively cover transient interactions. Therefore, overlaying and integration of predictive methods with experimental results provides statistical robustness and biological significance to the PPI data. In this chapter, the authors describe PPIs in terms of types, importance, and biological consequences. This chapter also provides a comprehensive description on various computational approaches for PPI prediction. Prediction of PPI can be done through: 1) Genomic information based methods 2) Structure based methods 3) Network topology based methods: 4) Literature and data mining based methods 5) Machine learning methods. For ease of use and convenience, a summary of various databases and software for PPI prediction has been provided.
Chapter Preview
Top

Introduction

Proteins are essential macromolecules that perform diverse functions in association with other macro and small molecules. Biological processes involve interplay between protein-protein interactions (PPIs) as cell to cell interactions, metabolic and developmental processes are governed by PPIs. A perturbation in the balance of PPIs occurs in various pathophysiological processes due to varying expression levels of disease related proteins. Specifically, PPIs also modulate the host-pathogen interactions in the initial stages and host response in later stages of infection. Modeling PPIs at a large scale for study of topological parameters and essentiality leads to identification of crucial players and pathways in the disease condition. Since proteins may have different interacting partners under diseased and normal state respectively, identification of drug targets from PPI networks is also carried out. Targeting PPIs presents a novel approach for developing therapeutics with high specificity and fewer side effects. Moreover, PPIs aid in forecasting the function of a target protein and drugability of a molecule. There are a large numbers of experimental methods for detection of PPIs that have led to the accumulation of PPI data in publicly available databases. However, experimental PPI determination is time consuming, expensive, error prone and do not effectively cover transient interactions. Therefore, prediction of PPIs by computational methods represents an important development. Overlaying and integration of predictive methods with experimental results provides statistical robustness and biological significance to the PPI data.

The chapter provides an overview on PPIs, its types, importance, biological consequences and factors affecting PPIs. The authors have also provided a comprehensive description on various computational approaches for PPI prediction. Computational methods for PPI prediction are based on sequence, structure, evolution and genomic data. Prediction of PPI can be done through:

  • 1.

    Genomic information based methods,

  • 2.

    Structure based methods,

  • 3.

    Network topology based methods,

  • 4.

    Machine learning methods and

  • 5.

    Literature and data mining based methods.

Complete Chapter List

Search this Book:
Reset