Evolutionary Computation in Gene Regulatory Network Research

CHAPTER 2MATHEMATICAL MODELS AND COMPUTATIONAL METHODS FOR INFERENCE OF GENETIC NETWORKS

Tatsuya Akutsu

Bioinformatics Center, Institute for Chemical Research, Kyoto University, Uji, Japan

2.1 INTRODUCTION

Genes maintain organisms by interacting with each other through messenger ribonucleic acids (mRNAs), proteins, and other types of molecules. Interactions among genes are often represented as networks, which are called gene regulatory networks, or genetic networks in short. Genetic networks are usually represented as directed graphs, in which nodes correspond to genes and edges correspond to regulatory relationships between two genes.

Deciphering genetic networks is important for understanding complex cellular systems because they play important roles in cells through control of protein production. In order to infer genetic networks, various kinds of data have been used such as gene expression profiles (particularly mRNA expression profiles), CHromatin ImmunoPrecipitation-chip data for transcription binding information, deoxyribonucleic acid (DNA)–protein interaction data, and mRNA seq data generated by using the next-generation DNA sequencing technology. In this chapter, we focus on inference of genetic networks from gene expression time series data (see Figure 2.1) because many studies have been done based on this kind of data and many of the developed methods and techniques may be applied to mRNA seq data.

Figure 2.1 Inference of a genetic network from gene expression ...