Modeling signaling networks using high-throughput phospho-proteomics

Abstract

Cellular communication and information processing is performed by complex, dynamic, and context specific signaling networks. Mathematical modeling is a very useful tool to make sense of this complexity. Building a model relies on two main ingredients: data and an adequate model formalism. In the case of signaling networks, we build mainly upon data at the proteome level, in particular about the phosphorylation of proteins. In this chapter we review recent developments in both data acquisition and computational analysis. We describe two approaches, antibody based technologies and mass spectrometry (MS), along with their main features and limitations. We then go on to describe some model formalisms that have been applied to such high-throughput phospho-proteomics data sets. We consider a variety of formalisms from clustering and data mining approaches to differential equation-based mechanistic models, rule-based, and logic based models, and on through Bayesian network inference and linear regressions. © 2012 Springer Science+Business Media, LLC.