Life in a Net

Shifting gears from the previous chapters, we describe a very different way to represent biological systems computationally: that of networks. We start with network examples from an ecosystem, the nervous system and the cell regulatory network. We then gain some intuition about networks by playing the Kevin Bacon game, and very briefly discuss the history of graphs and networks in mathematics and computation. A key insight is that many real networks have “small-world” properties, where the network is easily traversable in a few links but still preserves neighbor relationships. We show that both metabolic and protein interactions networks in real organisms possess these small-world characteristics and then discover a surprising relationship between these metabolic and protein–protein interactions. We next consider some directed networks where one entity operates on another, focusing first on neural networks and their recent use in artificial intelligence models. Moving on to transcriptional regulatory networks, we identify a circuit of transcription factors often seen in such networks. Reaching back to the model of noisy gene expression from Chapter 7 , we find that this circuit can reduce such noise, potentially explaining its prevalence.