Inference of Gene Regulatory Networks based on Association Rules

Abstract

This chapter focuses on gene regulation and the ways that transcriptome data can be used to unravel the complex relationships between the genes that comprise a gene regulatory network (GRN). In particular, it describes the main topics that must be considered in the field of association rule (AR) mining for reverse engineering of GRNs and presents the state-of-the art techniques currently available in the literature. The chapter presents the central concepts about AR mining for GRN reconstruction together with various other relevant issues. It reviews different data-mining approaches used for AR inference. The data-mining approaches are frequent-itemset-based methods, classification and regression tree-based approaches, Bayesian Networks, and Boolean networks. Over the past few years, other approaches were proposed that do not correspond to the previous classification. However, these methods can also be used to infer gene ARs from microarray data. These techniques are clustering, pairwise methods, and support vector machine methods.